The Human Signal: Where Intelligence Begins

AI becomes useful not simply when it knows more, but when it understands what matters. Through stories and engineering dilemmas, this keynote explores how trusted knowledge, business context and human judgment turn capable models into meaningful systems and why people remain essential for setting intent, recognizing nuance and defining what “good” looks like.

From Vibe Coding to Verified Coding: Shipping Production Web3 with AI

Vibe coding works for marketing sites. It fails for blockchain projects. The cost of "almost right" code scales with what your code touches — for a landing page it's a re-render, for an on-chain transaction it's irreversible.

At LimeChain we built LaiM internally to make AI-assisted development predictable enough to ship blockchain products. It's a structured methodology:with quality gates between each phase and a verification chain that goes deeper than tests. Blockchain-specific invariants — receipt confirmation before success UI, BigInt for amounts, address comparison via library helpers, discriminated union connection state — are encoded as enforced rules rather than hopeful guidelines. This talk covers the methodology, the principles behind it, and the design decisions we got wrong before we got them right.

When AI Changes Faster Than the Enterprise

Agentic AI is evolving faster than most organizations can procure, integrate, govern, and adopt it. Models, development frameworks, pricing structures, regulations, operating models, and customer expectations all change at different speeds. As a result, an AI solution can succeed technically and still become too expensive, too fragile, or strategically obsolete before it reaches scale.

Drawing on practical experience with enterprise AI initiatives, this session explores how to design systems for continuous change rather than for today’s preferred model or platform. It will examine architecture reversibility, vendor and model portability, cost volatility, governance, and the often-overlooked limits imposed by organizational and customer adoption. Attendees will leave with practical principles for building agentic AI capabilities that can evolve without forcing the enterprise to start over every time the technology changes.

Fewer Tokens, Faster UIs: Building Web Apps with Ignite UI, MCP and Agent Skills

We build the same data-rich web app UI three ways: in-house, with Ignite UI alone, and with Ignite UI plus its MCP servers and agent Skills. Then we compare developer time, license fees and AI token costs side by side.

One model reads your repo, takes orders from strangers, and hunts for holes in your code

The agent in your IDE has the whole repo, the terminal, and whatever you left in .env. For an single function, it might read forty files, and all of them leave your machine. Then the agent opens a README from a vendored dependency and does what the README tells it to do, without asking you first. The trust boundary moved while nobody was looking. It used to be user input. It is now every byte the model ingests, which includes your issue tracker, your logs, and a markdown file written by a stranger in 2019.

Meanwhile, roughly the same model, on someone else's laptop, is looking for holes in what you already shipped. Models like it have surfaced tens of thousands of previously unknown vulnerabilities in open source this year. Of those attributed to AI-assisted discovery, a little over a thousand were cataloged, and just 14 have been confirmed exploited in the wild. So the End of the World is postponed (again). Your backlog simply grew by an order of magnitude, the number of people reading it did not, and the median gap between a CVE going public and someone using it has fallen from 120 days to 80.

Don't expect slide defining what artificial intelligence is. No legal analysis. No FUD. Three places where AI is at once your most productive teammate, your least reviewed dependency, and your worst adversary, as well as what to actually do about each one.

The Future Starts at School: Education, Technology & Skills

Panel Description:

Artificial intelligence, automation and rapidly changing industries are reshaping the skills young people will need to thrive. While employers increasingly seek critical thinking, collaboration, adaptability, creativity, and digital literacy, education systems around the world are still working to bridge the gap between today's classrooms and tomorrow's workforce.

But technology alone will not prepare students for the future. Teachers are the single most important factor in helping young people develop the knowledge, mindset and competencies they will need to succeed. It is teachers who transform technology into meaningful learning experiences, inspire curiosity, encourage problem-solving and help students build the confidence to navigate an uncertain future.

This panel brings together perspectives from business, education, and the nonprofit sector to explore one essential question: What should we teach today to prepare students for the world students will enter tomorrow - and what support do teachers need to make this possible?

The discussion will examine emerging labour market trends, reflect on the implications of the latest PISA results (to be released in September) and showcase how technology can transform learning when teachers are empowered to innovate:

A Teach for Bulgaria alumni STEM teacher will present a real classroom project demonstrating how technology enables students to solve authentic problems while developing future-ready competencies.
The business perspective will highlight the skills employers increasingly value and why investing in education is an investment in the future workforce.
Teach For Bulgaria will connect these perspectives by sharing how developing and supporting outstanding teachers can help bridge the gap between today's classrooms and tomorrow's labour market.

Rather than focusing on technology for its own sake, the panel will demonstrate that the future of education is not about replacing teachers with technology, but about empowering excellent teachers to use technology in ways that help every student reach their full potential.

Ultimately, the discussion will show that preparing children for the future is a shared responsibility - one that requires collaboration between educators, businesses, families, and civil society.

Workshop: Human Systems Under Pressure: performance, recovery and tools for healthier work

Places are limited — reserve yours here

High performance is sustainable only when recovery becomes part of the system

In an environment where high performance is often the norm and constant connectivity blurs the boundaries between work and personal time, the most valuable resource is not time — but human energy.

This workshop offers a practical framework for recognizing and understanding the signs of prolonged workplace stress — from fatigue, lack of engagement and declining motivation to apathy and exhaustion — and how it affects concentration, motivation, decision-making and overall performance.

We will explore the early warning signs we often overlook — from chronic fatigue and reduced engagement to feelings of emotional exhaustion — and discuss how to distinguish natural periods of high workload from patterns that may pose a risk to our long-term well-being.

Participants will learn how to interpret these experiences in context, distinguish temporary pressure from longer-term trends, and consider what they may indicate about personal needs, workload and the broader work environment.

Through practical exercises and discussion, participants will explore how to manage their own energy more consciously, build healthier work habits, and create the conditions for sustainable performance without the cost of chronic exhaustion.

The session is designed for anyone who wants not simply to work more, but to work smarter, healthier, and with a focus on long-term effectiveness.

Traditional UI automation is highly effective at executing workflows and validating deterministic outcomes, but some categories of verification remain difficult to handle and maintain through conventional assertions alone. Scenarios involving dynamic content, semantic correctness, contextual UX expectations, and generated text often require validation beyond selectors, snapshots, and exact string matching.

To bridge this gap, Large Language Models (LLMs) can be integrated into Playwright automation as an evaluation layer—serving not as a replacement for deterministic testing, but as a powerful complement to it. This approach centers on building a custom "LLM-as-a-Judge" workflow capable of evaluating contextual UI behavior, semantic text correctness, and multimodal application output directly inside end-to-end tests.

Moving beyond the hype of "fully autonomous" AI testing, this session examines the practical engineering tradeoffs required to make probabilistic evaluation reliable enough for real QA pipelines. We will break down exactly where conventional assertions fail and demonstrate how to responsibly implement LLM evaluation using reproducibility techniques, confidence scoring, and robustness-testing techniques designed to verify test effectiveness against intentional UI changes.

Software-Defined Defence: Engineering the Next Generation of Defence Systems

"In August 2026, four drones reached European critical infrastructure - a NATO logistics hub, a gas pipeline, a runway, an offshore gas field. Three of the four were found by a human being rather than by a sensor.

Nothing that was missing on those mornings was made of metal. Detection, sensor fusion, classification and command-and-control are software problems, and they now decide whether a defence system works.

This session is an engineer's tour of software-defined defence. What the term actually means, once you strip the marketing off it. Why an unmanned vehicle is its purest case - remove the pilot and every function the pilot performed becomes code or an interface. And what the reference architecture looks like layer by layer: sensors at the physical edge, compute and communications above them, and the fabric that makes them interoperable.

Europe is stepping up. Investments, coalitions and programmes are moving faster than at any point in a generation, and for the first time the doors are open to software companies and not only to the primes. Bulgaria has one of the strongest software industries in Europe, with more than sixty thousand engineers in it. Contributing to this work is our opportunity - and, increasingly, our responsibility. This session is about where to start."

The Dark Side of AI Agents: Live Hacking Modern AI Applications

Description

https://github.com/DojoBits/ista-2026

Generative AI has evolved far beyond chatbots. Today's AI agents can execute code, call external tools, access sensitive data, interact with Kubernetes clusters, and make decisions with minimal human intervention. As AI capabilities evolve, so does the attack surface, forcing security to keep pace.

Through a live hacking demo, we'll show how a trusted AI agent can be manipulated into performing unintended actions without exploiting a traditional software vulnerability. Using the attack as a foundation, we'll explore how AI security is evolving through the latest work of the OWASP GenAI Security Project, examining why securing AI is no longer just about prompt injection but also about agent autonomy, tool integrations, and trust boundaries. We'll conclude with practical security patterns, including least agency, runtime guardrails, scoped identities, and sandboxing, that can be applied to real-world AI systems.

You'll leave with a practical understanding of the evolving AI threat landscape and concrete techniques for building and securing AI-powered applications.

What attendees will learn

Understand how AI security is evolving from LLMs to autonomous AI agents
See a live attack against a modern AI agent and understand why it succeeds
Learn how agent autonomy, tools, and trust boundaries create new security risks
Explore the latest guidance from the OWASP GenAI Security Project and the OWASP Top 10 for Agentic Applications
Apply practical security patterns to build more secure AI-powered applications

1,2,3 … testing : is this thing on(line)? Meet your Microsoft Testing tools

Is your environment acting the way you intended it to be, as in do your users see what you wanted them to see?

Is your app breaking under stress or even worse going down when components are acting up (or down in this case)?

In the past people were using Azure Devops Load Testing and related. But we all know some of these services have been deprecated. In this session you will be guided though all the options you have today lining out all the testing capabilities you have in the Microsoft Coding Universe.

Let’s take a stroll through the various options for load, chaos and automated testing in all things Microsoft devops and Azure. In doing so you will get to learn which services to use to improve reliability, performance usability and resilience of the applications you are building.

Note: testing includes Chaos Studio, Playwright, Azure Load Testing and more

AI-Powered Approaches to Legacy Modernization

Legacy systems don't fail overnight — they accumulate debt quietly. Decade-old codebases, undocumented dependencies, and untested critical paths are the norm across most enterprises. The real challenge isn't acknowledging the problem, it's knowing how to modernize without repeating the costly mistakes of full rewrites that delivered little to no business value.

This session focuses on what actually works in AI-assisted modernization. We'll walk through how to frame the rewrite versus incremental refactoring decision honestly — accounting for risk, team capacity, and business-critical constraints. You’ll learn where AI tooling earns its place, like understanding unfamiliar codebases, generating test coverage, and surfacing dependency hotspots before they become bottlenecks.We'll also cover the failure modes and why AI output without a human review loop is just technical debt wearing a different label.

You'll leave with a practical overview of the current tooling landscape and a set of questions worth asking before any migration begins.

The Human-in-the-Loop Engineering Organization

AI is becoming an active participant across the product development lifecycle — from product discovery, design, and planning to software development, testing, documentation, and management decision-making. While this raises concerns about AI replacing people, the more immediate and important shift is that AI will change how product teams work, collaborate, and make decisions. This session explores the idea of the Human-in-the-Loop Engineering Organization: a model where AI accelerates work across roles, but humans remain responsible for intent, context, creativity, judgment, quality, and outcomes. We will look at how designers, product managers, engineers, QA professionals, and leaders can redesign workflows so AI becomes a trusted collaborator rather than an uncontrolled shortcut.

In this session, Mihail will share industry observations, practical insights, and lessons learned from Progress Software’s AI adoption journey to show how product development organizations can increase speed and capability while preserving human ownership, accountability, and product thinking.

How and Why Deepfake Threats are Evolving for Dark-web Criminals

In this presentation Thomas ‘Mannie’ Willkan explores how and why deepfakes are being utilized by dark-web criminals to target enterprises. The presentation explores TTPs and motivations associated with malicious deepfake use and moves far beyond merely examining publicly known attacks. The work is based on years of Dark-Web Recon research and real life incidence response cases as well as novel insight into how the groups weaponize leaked enterprise data and what most defenders overlook.

JavaScript SEO in the Age of AI Crawlers

"As search evolves beyond traditional indexing into an era dominated by AI powered crawlers, JavaScript-heavy websites face new opportunities - and new risks. This session explores how modern AI search systems process, interpret, and evaluate JavaScript-rendered content, and what SEO professionals, developers, and product teams must do to stay competitive.

Attendees will learn how AI crawlers differ from legacy bots, what signals they prioritize, and how JavaScript frameworks can either help or hinder AI-driven understanding. The lecture will cover real-world examples of rendering pitfalls, indexing delays, and content misinterpretation, as well as practical solutions such as dynamic rendering, component-level SEO best practices, and semantic optimization for AI models.

Finally, the session will outline how AI-generated search results (AI Overviews, answer engines, and generative snippets) reshape traffic patterns - and how teams can adapt their JavaScript architecture and content strategy for maximum visibility in this new landscape.

Key Topics Covered

• How AI crawlers read and interpret JavaScript

Including differences from traditional HTML-first crawling and processing pipelines.

• Concrete recommendations for dev + SEO collaboration

Developers lead the workflow and decide what’s feasible, while SEOs support by adding clarity and structure to strengthen AI visibility - making seamless collaboration between both sides essential.

• JavaScript rendering challenges in AI-driven search

Common issues with hydration, lazy-loading, dynamic routing, and client-side content injection.

• Technical SEO strategies for modern JS frameworks

The benefits of server-side rendering, optimising JavaScript delivery while avoiding both excessive fragmentation and large bundles.

• Preparing content for AI-driven search experiences

How to structure information for AI summarization, answer extraction, and model-level relevance scoring.

Benefits for Attendees

• Practical, immediately actionable SEO techniques tailored for JavaScript environments.

• A clear understanding of how AI crawlers actually “think”, enabling smarter architectural decisions.

• Improved collaboration between SEO, dev, and product teams, supported by technical examples and checklists.

• Future-proofing strategies to maintain visibility as search continues shifting toward generative AI answers."

One does not simply add AI to their apps: stop and Think 1st

Everyone wants to build AI infused applications, but if you want to do this you need to have your foundations solid first. And this is where it typically goes wrong. A lot of solutions aren't set up for this. This session will give you solid architectural design principles to make sure you can get started with AI. The principles will be demonstrated and discussed on Microsoft technology and Azure but can be adapted to any other cloud or tech.

Who Should Own Your Customers’ Analytics? What Reveal and Claude Dashboards each do best

Most data never gets looked at. It sits in tables, reports and exports, and the people who could act on it can’t see what it’s telling them. That’s changing quickly: today you can ask a question in plain language and watch a live dashboard build itself in seconds.

Anthropic’s Claude Dashboards, launched in beta this month, does exactly that. Anthropic says it works well for “quick, exploratory questions.” And then there’s a second case worth looking at: when the person asking is your customer, working inside your product and looking at their own data.

That’s the problem software companies have to solve. They ship analytics to thousands of end users, under their own login, and each customer must see only their own data, which often lives on infrastructure it can’t leave. This session walks through both Claude Dashboards and Reveal as two working examples of how analytics is built today. You will see how AI generates dashboards from plain-language questions in each of them.

We close with a practical guide to what each approach does best and when to reach for it. We also share a lesson from years of building alongside platform vendors: sooner or later every platform ships a basic version of your feature, and you win by owning the layer it can’t reach.

What attendees will learn

How AI is turning plain-language questions into dashboards anyone can read, and where that works today
What a chat-based dashboard tool is great at, and what in-product analytics adds on top
How to embed white-labeled analytics in an existing app
How to control on the server what each tenant sees
How to let AI build dashboards with the model of your choice, from Claude to a local LLM

[PANEL] Burnout: fact, fiction and the cost of success. The paradox of productivity

Burnout is increasingly present in conversations about mental health and workplace performance, particularly in dynamic corporate and technology environments. More and more professionals are talking about exhaustion, lack of motivation, and a loss of meaning in their work. But is burnout truly a condition caused by chronic stress and accumulated exhaustion, or has it become an overused, broad term for dissatisfaction, fatigue, and loss of motivation?

This panel will not provide a simple answer — instead, it will challenge established perceptions. Psychologists and representatives from the IT industry will come together for an open debate to explore whether burnout is an inevitable cost of high performance, the result of toxic organizational culture and chronic stress, or simply an overused term we apply to every experience of professional frustration, exhaustion, or lack of motivation.

Burnout is a real syndrome recognized by the WHO, costing businesses billions and potentially having a devastating impact on people’s mental and physical health. In today’s fast-paced corporate world, success and stress often go hand in hand — but where does healthy pressure end?

We will discuss the early warning signs we often overlook, the cost of constant connectivity, the role of leaders and organizations, as well as personal responsibility, the ability to set boundaries, and the importance of managing our own energy.

Because not every kind of fatigue is burnout — but not every case of burnout begins with the feeling that we have reached our limit.

Where does healthy pressure end and true exhaustion begin? And can we build a culture of high performance without it coming at the expense of people?

End-to-End Control: Governing Autonomous AI Agents

Modern AI agents powered by large language models can plan, use tools, remember past interactions, and complete complex workflows with minimal human input. That's powerful — but it also creates new risks. Agents can hallucinate. They can leak sensitive data. They can take unintended actions. They can be manipulated through prompt injection or drift away from their original goals.

The key question is: how do we stay in control while letting agents remain autonomous?

This talk introduces the End-to-End Control architecture — a runtime supervision and governance layer designed specifically for agent systems. The framework was developed and battle-tested through real-world implementations deployed across multiple production systems at WebPros, where we share our experience of building such a system for our zoo of projects using LLMs. Instead of only focusing on model training or prompt engineering, this approach emphasizes continuous control during execution.

We will cover 4 layers in this presentation:

Observability
Governance
Evaluation

We'll explore:

Why runtime control is essential for autonomous systems

The difference between protecting inputs (e.g., defending against prompt injection and context poisoning) and validating outputs (e.g., checking actions before they're executed)

How to build deep observability into agent systems — including tracing reasoning steps, monitoring tool usage, and tracking memory updates

How to enforce security policies in real time without breaking the agent's workflow

The goal is simple: enable powerful AI agents without giving up safety, transparency, or control — proven in production at WebPros.

The Invisible Architecture: Why Your Testing System Needs a Constitution

Every testing team runs on unwritten rules - which selectors to use, why that wait exists, what design patterns we use, where test cases belong. This tribal knowledge governs every line of test code, and because it lives only in heads, frameworks drift. AI makes it worse: an agent has none of that knowledge, so it drifts at the speed of generation.

The fix takes three moves. Capture the rules in a testing constitution — MUST, SHOULD, WON'T. Deliver them through context engineering, so the agent gets exactly the knowledge each task needs - never a book it can't hold. Enforce them through loop engineering - explore, plan, verify, with hard gates the agent can't talk its way past. Same request, same model: several violations ungoverned, zero governed. You'll leave a clear mental model why you need this.

Workshop: Hack the Agent: Hands-on AI Agent Security

https://github.com/DojoBits/ista-2026

AI agents are rapidly becoming part of modern applications, capable of executing code, calling external tools, accessing sensitive data, and interacting with cloud-native platforms. While these capabilities unlock new possibilities, they also introduce an entirely new attack surface where traditional application security falls short.

This hands-on workshop builds on the conference talk and focuses on ASI02: Tool Misuse and Exploitation from the OWASP Top 10 for Agentic Applications. Participants will exploit a vulnerable AI agent by manipulating its interaction with external tools, then analyze why the attack succeeds and progressively harden the application using practical security controls such as least agency, scoped identities, runtime guardrails, and sandboxing.

By the end of the workshop, you'll have hands-on experience exploiting and defending AI agents, along with practical techniques you can immediately apply to build more secure AI-powered applications.

What attendees will learn

Exploit a vulnerable AI agent using real-world attack techniques
Understand why tool misuse is one of the most critical risks for AI agents
Analyze how agent autonomy, tools, and trust boundaries contribute to security failures
Implement practical defenses including least agency, runtime guardrails, scoped identities, and sandboxing
Apply guidance from the OWASP GenAI Security Project and the OWASP Top 10 for Agentic Applications

Requirements

Bring your laptop.
Every participant will receive a dedicated, isolated lab environment. No cloud account or local setup is required.
An SSH client is required. If you don't have one installed, we'll help you set it up before the workshop.
We'll be available throughout the workshop to answer questions and help if you get stuck.

Building AI-Ready Organizations: Culture, Skills & Change Management

AI adoption isn’t just a tech rollout — it’s an organisational shift.

Companies often buy powerful tools but overlook the human side. Culture, trust, learning speed, and leadership behaviour decide whether AI boosts performance or becomes an expensive experiment.

This session shows what makes an organisation truly AI-ready. Blending neuroscience, leadership psychology, and real change practices, it explains how to reduce fear, speed up adaptation, and build curiosity instead of resistance.

You’ll leave with a practical roadmap to align culture, skills, and leadership so AI strengthens engagement and innovation — without losing the human core.

Levelling Up: From Diablo II Scripts to Automated Security Audits

Remember the adrenaline of chasing that elusive Shako in Diablo II — the non stop grinding and the sweet victory of a perfect drop? Gamers mastered strategies to farm treasures faster than ever, and now it’s time to bring that same energy to QA. This session will demonstrate how idea behind Diablo II bots can be used nowadays, turning predictable environments and repetitive tasks into streamlined, error-free workflows. Why settle for manual scans when you can script your way to legendary results? Learn how to transform QA grind into epic wins, crush human error, and unlock efficiency like a true endgame boss. Ready to claim your ultimate treasure chest of automation? Join us and level up your QA game!

Quantum Investments: from Fundamental Science to Strategic Advantage

"Investment in quantum technology grew 6.3 times in a single year. That number is real - and three other credible trackers published figures for the same year that were a third of its size. Understanding why is the fastest route to reading this field honestly.

This session takes an engineer's view of the quantum market. We look at what actually changed in December 2024, when error correction crossed the threshold in a peer-reviewed result and fault tolerance stopped being a physics question. We look at why key market analysts are five years apart on when it becomes commercially useful, and how to tell a peer-reviewed result from a vendor announcement. The session will also cover where the investments are going, why 94% of Chinese quantum patents never leave China, and where Europe is genuinely strong.

What quantum needs next is mostly software - control stacks, error-correction decoders, compilers, HPC integration - and the scarcest resource in the field is engineers and physicists. With one of the strongest engineering communities in Europe, Bulgaria should be aiming to build one of the strongest quantum capabilities on the continent."

AI-accelerated Smart Product Development

AI is rapidly changing the way software and hardware products are designed, developed, tested, and brought to market. In the world of IoT and smart devices, this opens entirely new opportunities for manufacturers to prototype faster, automate development workflows, and build more flexible products.

This talk explores how Shelly X enables manufacturers to create their own smart products on top of the Shelly ecosystem, while leveraging AI agents throughout the product lifecycle.

Rather than focusing on theory, the presentation will show how AI is already becoming a practical engineering tool for modern IoT development and how platforms like Shelly X help turn ideas into production-ready smart products much faster than before.

AI: The Line Connecting the Dots - From Test Cases to a Product That Builds Itself

Every one of us carries a set of skills and experiences that can feel disconnected — a degree, a hobby, a job we assumed was temporary, a chapter of life we figured would never matter again professionally.

Years spent teaching people switching careers into Quality Assurance and Automation taught me something I didn't expect: it's rarely just the tech skills or the testing theory that makes someone good at this job. The best QA engineers are consistently the ones testing systems closest to their own past experience and domain knowledge — the dots they already had, showing up exactly where they were needed.

I kept adding dots of my own for years — an Electrical Engineering background, a decade in QA, a habit of automating my own home, a business education. None of them had a reason to connect. Then I found a project ambitious enough to need all of them at once, and discovered that AI finally gave me a real chance to connect them.

This talk tells that story, step by step, including something I'll build live, on stage, in real time. It isn't a talk about what AI can do. It's a talk about what happens the moment it lets you build with everything you actually are.

Walk away with a simple challenge: find the dots you've been carrying that have never had a reason to connect. Then find the project ambitious enough to need all of them at once.

The Hallucination of Trust: Why AI Breaks Teams and How to Reclaim Ownership

AI is making teams faster, more productive, and increasingly automated. But beneath the surface of green performance dashboards and optimised workflows, a silent culture crisis is brewing.

In this session, executive coach and business strategist Dimi Ilieva re-examines Patrick Lencioni’s classic The Five Dysfunctions of a Team through an AI-era lens. We’ll explore how AI is changing the way five familiar dysfunctions show up inside modern teams.

And then we’ll take the idea one step further: what happens when the team itself is made up of autonomous AI agents?

Walk away with a practical framework to lead beyond the software, protect team culture, and reclaim true decision ownership in an AI-driven workplace.

Panel discussion: Tech at a Turning Point: The Future of IT in Bulgaria

TBA

Ayca Bas

Ayca Bas is a Developer Advocate at Microsoft, working at the intersection of product, engineering, and developer communities to drive the practical adoption of AI. She designs and scales developer programs, hands-on experiences, and content frameworks that turn complex capabilities into structured and applied learning journeys. Her work focuses on building repeatable programs that enable developers to go from exploration to implementation at scale. She regularly shares insights and learnings with the broader developer ecosystem on LinkedIn.

Ognyan Chikov

Ognyan Chikov is Head of Engineering at LimeChain, with a background in software engineering, blockchain, and distributed systems. He has worked on Web3 projects across decentralized finance, cross-chain infrastructure, and zero-knowledge technologies. In his day-to-day work, Ognyan focuses on scalable and secure blockchain systems, interoperability, performance, and leading the company's internal AI software development lifecycle (AI SDLC) integrations to enhance engineering productivity and development workflows. He has also mentored developers through LimeAcademy, helping engineers transition into the Web3 space and gain a deeper understanding of the infrastructure behind decentralized systems.

Iliyan Nenov

Iliyan Nenov works at the intersection of business strategy, solution architecture, and AI adoption, helping organizations move from ambitious ideas to practical technology decisions. In his current role, he advises international companies on AI implementation, technology roadmaps, and the real-world challenges of making innovation work beyond prototypes and presentations.

Over more than two decades, his career has moved through software engineering, product management, platform strategy, engineering leadership, and business development. This has given him a broad, hands-on view of how enterprise technology is built, adopted, scaled, and sometimes misunderstood.

He also brings an academic perspective, with a PhD in Economics focused on applying ML and AI approaches to complex economic forecasting. His recent research explores how advanced AI techniques can complement classical econometric methods.

His talks usually focus on the gap between AI expectations and AI reality: what is genuinely useful, what often gets overlooked, and what organizations need to consider if they want AI initiatives to create lasting value

Konstantin Dinev

Konstantin Dinev is Director of Product Development at Infragistics, with more than 15 years of experience in product engineering and technical leadership. He leads the development and evolution of enterprise software products and drives the adoption of modern, scalable software development practices across the organization. A recognized industry expert, Konstantin has spoken at conferences such as Visual Studio Live!, Enterprise NG, and Angular Connect, and has appeared on The Angular Plus Show and the Azure DevOps Podcast. He regularly shares technical insights through articles published on Medium, Infragistics blog, App Builder blog, The CTO Club, Solutions Review, Top End Devs, and has served as a board member of the ISTA Conference since 2015.

Peter Kirkov

Peter Kirkov is an experienced cybersecurity consultant with over 20 years of experience in strategic management, regulatory compliance, and the practical protection of information systems. His expertise in regulatory frameworks such as NIS2, GDPR, DORA, PCI-DSS, and others positions him as a trusted partner for organizations navigating the growing risks associated with AI-based infrastructure.

From 2021 to 2024, following his extensive experience in the private sector, he served as National Cybersecurity Coordinator of Bulgaria and Director at the Ministry of E-Government. In these roles, he led national-level incident coordination, international cooperation with the EU CSIRT network and CyCLONe, and the transposition of the NIS2 directive into national legislation. He represented Bulgaria as an alternate member on the governing boards of European agencies and led projects to establish national-level SOC capabilities.

Prior to his roles in public administration, his career spanned IBM, Deloitte, and Telelink Business Services, with a focus on IT audit, risk management, and policy development and compliance program implementation.

He holds ISACA CISA, ISO/IEC 27001 Lead Auditor, and ISO/IEC 42001 Lead Auditor certifications. He has delivered training at the University and High school levels and is a frequent keynote speaker at public cybersecurity events.

Trayan Trayanov

Trayan Trayanov is the Executive Director of the Teach for Bulgaria foundation - a non-profit organization working to ensure equitable access to quality education for every child in Bulgaria. Trayan was one of the first people to join Teach for Bulgaria during its establishment in 2010, where he was responsible for the admissions, selection, teacher training and the monitoring and impact of its programs. He assumed the role of a CEO in 2019 and now manages the strategic development and sustainability of the organization.

Trayan holds a BA in Social Studies from the University of Harvard and a MA in Sociology from the University of Cambridge. Before joining Teach for Bulgaria he worked as a business analyst for the Mitchell Madison Group in New York and London and for Arthur D. Little in Prague.

Trayan is a long-time advocate for educational equity, driven by questions such as: How should schools and classrooms be organized to offset the effects of the ‘lottery of birth’ or the misfortune of being born and raised in a disadvantaged environment?

Trayan is also on the Board of Trustees of the American College of Sofia where he completed his secondary education.

Georgi Stoev

Georgi Stoev holds a degree in architecture from the University of Architecture, Civil Engineering and Geodesy in Sofia. He discovered his calling as a teacher in 2017, when he joined Teach For Bulgaria’s New Way to Teaching program, and today teaches mathematics at Lyuben Karavelov Secondary School in Plovdiv. In his work, he uses mathematics not only as an academic subject, but as a tool for developing students’ critical thinking, global competence, and civic skills.

Georgi designs educational practices that connect mathematics with real-world social challenges, including projects on designing a refugee camp for children in Africa and planning the post-war reconstruction of a children’s hospital in Gaza. His work has been recognized by initiatives including Global School and Ideas That Change the World at Sofia University. He has also worked on topics related to children’s rights and youth participation in partnership with national and European organizations.

Zlatina Georgieva

Zlatina Georgieva is an education consultant, teacher trainer and former teacher whose work focuses on the quality of teaching and learning, teacher development, and the role of AI and technology in education.

She entered the education sector through Teach for Bulgaria, where she first worked as a primary school teacher and later as a teacher trainer and mentor. Today, she works closely with schools and educators, observing classroom practice and supporting teachers in creating learning environments that foster student engagement, independence, and meaningful learning.

Before moving completely into education, Zlatina spent nearly 20 years in media as a journalist and editor, including more than 12 years at Forbes Bulgaria. This combination of media and education experience shapes her interest in how schools can respond to a rapidly changing world and prepare young people with the skills they will need beyond the classroom.

She is the founder of "365 идеи", an educational platform for teachers and parents. Her current work also includes training teachers on the practical and responsible use of artificial intelligence in education.

Natalia Miteva

Natalia Miteva is a Government Engagement and Education Policy Consultant with Google for Education (EMEA), leading national-scale AI initiatives and digital transformation strategies across the region.

A seasoned policymaker and executive, Natalia previously served as Bulgaria’s Deputy Minister of Education and Science (2023–2025). There, she oversaw a €1 billion education portfolio, spearheaded the nation's inaugural AI Guidelines in Education, led the education track for Bulgaria’s OECD accession, and established strategic multilateral partnerships with the World Bank, OECD, and the European Commission.

Her career bridges public policy, global philanthropy, and the startup ecosystem, having served as Program Director for Education at the America for Bulgaria Foundation, Country Lead for HundrED, and an advisor to high-growth EdTech ventures.

Ralitsa Mihaylova

I’m a clinical psychologist with nearly a decade of experience working across different settings with both children and adults. I strongly believe that taking care of our mental health is not something we should think about only when there is a problem, but an essential part of building a happier, healthier and more sustainable life. I have a particular interest in prevention and understanding human systems under pressure. In my work, I focus on how language shapes the way we understand ourselves and our experiences, and on cultivating the curiosity to explore how we relate to the world around us.

Hristina Damyanova

As a clinical psychologist, I believe psychological support is not only about overcoming difficulties, but also about developing greater self-awareness, resilience, and healthier ways of navigating life and change. In my work with children, adolescents, and adults, I focus on creating a safe space for curiosity, understanding, and meaningful change. I believe that understanding ourselves is the foundation for how we relate to others, respond to challenges, and navigate uncertainty. Meaningful change begins with understanding — and with creating the space to look at ourselves with curiosity rather than judgement

Atanas Ganchev

Atanas Ganchev is a clinical psychologist with qualifications in cognitive-behavioral psychotherapy and Mentalization-Based Treatment for Families and is currently training in psychoanalysis. His clinical work covers a broad range of emotional and relational difficulties, with a particular interest in stress, burnout, adaptation, and the ways psychological functioning is shaped by relationships and working environments. Alongside his clinical practice, Atanas leads workshops and seminars for professionals and organizations, develops and facilitates psychoeducational programs and activities, and has presented at national and international conferences on psychological development, mental health, and human relationships.

Anton Turbinov

Spending my early career in Customer Success, where I analyzed customer conversations gave me a sharp eye for natural interaction. I learned firsthand that all systems automated or not will only succeed when they are built around the nuances of real-world user behavior. Today, I leverage my 15-year IT background to build AI-native quality assurance systems from scratch. I spend my time engineering end-to-end automation workflows—using Playwright and TypeScript. Lately, I've been building "LLM as Judge" multimodal frameworks—essentially creating the evaluation layer where advanced models review and critique the work of other autonomous agents generating web and social content. I am a huge fan of the idea to make AI testing pragmatic and thorough so development teams have a clear, measurable data on how their AI is performing.

Dimitar Dimitrov

Dimitar Dimitrov is Co-Founder and Board Member of Wiser Technology (BSE: WISR), a publicly listed European engineering group of more than 500 engineers serving the high-tech, banking, telco, automotive, space and defence sectors. His twenty-five years in advanced software and systems engineering include VP of Research & Development at SAP, where he drove global technology-platform development, and Founder & CEO of Digital Lights, named an "Impact Star" in the Deloitte Technology Fast 50 Central Europe. He also builds and backs deep-tech ventures of his own, among them Bronia in acoustic sensing and AI for defence, and Team Schedule in enterprise AI and the future of work. Dimitar lectures at Sofia University's Faculty of Mathematics and Informatics and sits on the boards of BASSCOM, the German-Bulgarian Chamber of Industry and Commerce and the Bulgarian Space Cluster. He was named one of Bulgaria's "40 under 40" most influential IT leaders.

Valentin Hristev

Recognized by Harvard Business School, CNCF, AWS, VMware, and others, Valentin Hristev is a Cloud Native consultant and trainer with over two decades of experience. He specializes in keeping cloud environments clear and storm-free, helping companies implement cutting-edge solutions. Passionate about innovation and challenges, Valentin empowers individuals through continuous learning and actively participates in non-profit organizations and community events. He enjoys spending time with his dog and exploring his passion for cars. Valentin's dedication to technology and personal growth is complemented by his drive to master life's bumpy and uncertain courses.

Iliyan Petkov

Iliyan Petkov is a passionate technologist dedicated to creating innovative solutions that improve efficiency and make life better. With a proven track record of working with some of the world's leading IT companies, Iliyan has been recognized with the prestigious CNCF Kubestronaut designation. When not involved in training or working on projects, Iliyan enjoys activities such as reading, martial arts, programming, and learning new languages. He also loves spending time with his family, discussing technology at conferences and workshops, helping non-profit community initiatives, and tinkering with electronics and IoT, which fuel his creativity and curiosity.

Mike Martin

As a Technical Evangelist with a large passion for Microsoft Technologies, Mike is an Azure goto for many. He’s been active in the IT industry for 30 years and has performed almost all types of job profiles, going from coaching and leading a team to architecting and systems design and training. Today he’s primarily into the Microsoft Cloud Platform and Application Lifecycle Management. He’s not a stranger to both Dev and IT Pro topics, they even call him the perfect hybrid solution.

In January 2012 he became a crew member of AZUG, the Belgian Microsoft Azure User Group. As an active member he’s both involved in giving presentations and organizing events (like Cloudbrew, Techorama and Global Azure Bootcamp aka GAB). Mike was and is also a Microsoft Azure MVP (awarded 5 times since 2013, before going into Microsoft as a bluebadge and he regained the status after leaving in December 2025) and Microsoft Azure Advisor.

Helping out in the community and introducing new & young people into the world of Microsoft and technology is also one of his passions.

Yoanna Dimitrova

Yoanna Kostova is a Senior Software Consultant at Accedia with over 5 years of hands-on experience across full-stack development, cloud-native architecture, and AI-enabled systems. Her work spans financial services, logistics, media, automotive, and insurance – domains where legacy complexity is the rule.
Recently, she has taken on technical leadership roles in AI product development, guiding teams through the real challenges of integrating machine learning into production systems.
Yoanna is one of the technology consultants driving Accedia's AI capabilities — producing reusable AI models, spreading practical AI skills across the company, and empowering team members to use AI securely and ethically.

Mihail Vladov

Mihail Vladov serves as a Senior Manager of AI Operations at Progress Software, leading efforts to accelerate AI adoption and deliver practical, scalable AI solutions across teams. With over two decades of product development experience, 15 years of which at Progress, he bridges engineering and product strategy to help organizations turn emerging AI capabilities into measurable business impact.

Thomas ‘Mannie’ Willkan

Thomas ‘Mannie’ Willkan is Head of Research & HUMINT Intelligence and experienced cybersecurity practitioner working on the dark web reconnaissance team at Lab-1 and previously Accenture Cyber Intelligence. Mannie specializes in research on threat-actor doxxing, deepfake threats, extortion negotiation and emerging trends on the dark web. Mannie has presented his original research at Black Hat USA and Europe multiple times as well as various ISACs and SANS Security Summits. Mannie has a War Studies Master of Arts degree with a focus in science and security from Kings College London. Originally from Copenhagen, he resides in Denmark after having lived abroad in London and Japan.

Vasilisa Marinova-Kaseva

Vasilisa is a Senior SEO Manager at Merkle Bugaria with over 15 years of hands‑on experience in search engine optimisation. She specialises in complex, JavaScript‑driven architectures, including SPAs and large‑scale enterprise websites. Over the past four years at Merkle Bulgaria, she has led SEO strategy for major global accounts such as BAT, PMI, and Crocs. During the years, she oversaw SEO for the largest educational website in Bulgaria and its international counterparts. Vasilisa is a Winner of the Total Search Awards 2026, sponsored by Google, recognised for excellence in integrated search through the successful combination of SEO, technical optimisation, AI, PPC, content, and advanced data utilisation for Benefit Cosmetics

Stefani Sankelova

Stefani has six years of experience in SEO, with a strong focus on technical SEO. Her expertise includes crawlability, indexation, site architecture, internal linking, page speed, Core Web Vitals, rendering, structured data, and international SEO. Alongside technical delivery, she has contributed across content, on‑page optimisation, and broader organic search strategy. Over the last three years at Merkle, Stefani has worked on major international brands including Tesco Mobile, Jimmy Choo, Gucci, WWF, and SkyShowtime, supporting complex technical initiatives and driving sustainable organic growth.

Zdravko Kolev

Zdravko Kolev is Manager of Product Development at Infragistics, where he leads work on Reveal, embedded analytics SDK. His work focuses on developer tooling and on how AI is changing the way software products are built, shipped and used.

In more than 12 years at Infragistics, he spent years fully dedicated to App Builder, the company’s low-code app development platform, and to Ignite UI, its UI component library for web developers. That background in developer tooling, SDK product management and technical content now shapes his work on Reveal. Today he leads a globally distributed team spanning the US, Uruguay and Bulgaria.

He has also driven AI adoption across the organization through hands-on work. He builds MCP servers, RAG pipelines, governance scaffolding and agentic workflows for internal teams, and he designed the AI Skills architecture for the Reveal SDK. That experience shapes his ISTA session, which compares embedded analytics with AI-generated dashboards and asks who should own a product’s customer-facing analytics.

Zdravko has published more than 45 articles on platforms including Solutions Review and the App Builder blog. He is also active in the Bulgarian tech community as a mentor at AIBEST and a lecturer at 9Academy.

Aneliya Georgieva

Aneliya Georgieva is Head of Communications and Brand for Experian’s Global Centres, leading communications and brand strategy across key markets including India, Bulgaria and Malaysia. With more than 12 years of experience in corporate communications, PR, change management and stakeholder engagement, she has worked across complex, international organisations to build reputation, engage employees and support business transformation. In her current role, she partners with global senior business and HR leaders to drive strategic communications. Outside of work, her interests include sports and travelling.

Petar Torneff

TBA

Dmitrii Ushakov

Hi, I'm Dmitrii, a Senior Machine Learning Engineer in the Applied AI team at WebPros with 8+ years of experience across a variety of domains.
I specialize in building AI and machine learning systems from scratch, including ML algorithms, AI architectures, and production-ready MLOps pipelines, with recent work on LLM agent-based systems.
I also teach AI productization at Product School and have spoken at the Warsaw Data &AI Tech Summit.
Link to my speech: https://www.youtube.com/watch?v=TMT5zpVdNOo&t=304s

Ivan Davidov

As a former Civil Engineer, I learned early on that without a solid foundation, any system will eventually drift, crack, and collapse. Today, I apply those same architectural principles to software quality, helping QA teams build practical, AI-native Playwright testing systems with zero hype and robust TypeScript workflows. My focus is on building "invisible architecture" - processes and systems that actually work to bring confidence in the whole development team.

Ralitsa Dobreva

AI adoption isn’t just a tech rollout — it’s an organisational shift.

Companies often buy powerful tools but overlook the human side. Culture, trust, learning speed, and leadership behaviour decide whether AI boosts performance or becomes an expensive experiment.

This session shows what makes an organisation truly AI-ready. Blending neuroscience, leadership psychology, and real change practices, it explains how to reduce fear, speed up adaptation, and build curiosity instead of resistance.

You’ll leave with a practical roadmap to align culture, skills, and leadership so AI strengthens engagement and innovation — without losing the human core.

Nikolay Stanoev

Nikolay Stanoev has 17+ years of experience in the IT sector. Starting as an intern in 2009, his passion for quality propelled him to the role of QA Architect, where he’s built test automation for major brands like Adidas, Disney, and Shiseido in the eCommerce sector. Niki has also mastered various types of non-functional testing. His influence extends beyond his projects—he’s spoken at renowned conferences like Automation Guild, ISTA, TestCon Europe and QA Challenge Accepted.

Andrey Gruev

Driven by a deep passion for the future of connected hardware, Andrey Gruev is Senior Product Owner for ShellyX at Shelly Europe, where he directs the “Powered by Shelly” portfolio through partner-led product development, commercial negotiation, and technical integration. He brings over thirteen years of hardware and software experience, rooted in professional audio and now driving innovation in IoT. Previously, as Senior Product Manager at Antelope Audio, he introduced the company’s first subscription-based revenue stream, launched flagship hardware including Zen Go Synergy Core and Galaxy 64, and expanded key integrations like the afx2daw plugin bridge.

Nikolay Avramov

Nikolay Avramov has over a decade of experience as a Quality Assurance Engineer. His story, however, doesn't start with software - it starts with circuits. Graduating from a specialized Electronics school and earning a Bachelor's degree in Electrical Engineering from the Technical University of Sofia, he built his foundation on one uncompromising principle: systems must work reliably in the real world, not just in theory.

That mindset carried him into software quality, where he grew from Telerik Academy Trainee to Technical Director and Co-Owner of Automate The Planet - passing through every level of the QA discipline along the way. Today he embeds himself into organizations as an Automation Architect, helping teams transform their entire approach to testing and automation from the inside out. He has helped compress release cycles from once a week to three times a day, contributed to BELLATRIX - one of the most advanced open-source .NET testing frameworks - and continues to shape the next generation of engineers as a Lecturer at Telerik Academy and a speaker at international conferences.

Dimi Ilieva

Dimi Ilieva is an executive coach and business strategist with 20+ years of experience in international business and leadership, including more than six years as Chief Marketing Officer and Board Member at myPOS, a fintech company operating across 30+ markets. Today, she works with leaders, teams, and organisations on leadership, decision-making, and organisational effectiveness, combining hands-on executive experience with a coaching perspective. Dimi is an ICF-certified coach and holds an MBA, a second master's degree in Entrepreneurship, and executive leadership education from Harvard.

Ilia Krustev

Ilia Krastev is a seasoned leader in the knowledge-intensive and technology sector, with over 20 years of experience in building, scaling, and managing innovative companies, products, and services with international reach.

As Managing Partner of A Data Pro, Ilia has led the company’s transformation into a leading international provider of data solutions, media intelligence, and risk intelligence services for global clients across multiple markets.

Ilia also holds the role of Chairman of the Bulgarian Employers’ Association for Innovative Technologies (BRAIT) and is a Board Member of AIBEST, where he contributes to strategic dialogue between business, education, and government institutions. His work is focused on modernising education, strengthening the talent pipeline, and fostering a predictable investment environment that can position Bulgaria as a competitive regional technology hub.

Ilia Krastev holds an MBA from Cotrugli Business School, a Certificate in Global Management and Leadership from Ritsumeikan Asia Pacific University, and a specialisation in Human Resource Development in Small and Medium-Sized Enterprises from Tokai University.

Krum Hadjigeorgiev

TBA

Ivan Rusev

Seasoned professional in the fields of Technology, IT and Operations with diverse geographic experience (EMEA, Asia and Americas) across a wide range of products and services. Currently he is Head of Software Development for Experian (Data Quality) in Sofia, Bulgaria. His remit is delivering enterprise level software for Experian's customers. Ivan is also a board member of the largest Bulgaria IT association - AIBEST, where he heavily contributes to their education agenda.

Personal motto: Get the right things done.