AITEX Summit Summer 2026 wrapped its two day event from August 30 to August 31. Working under the theme “From Ideas to Impact,” developers, engineers, designers, data scientists and technical leaders formed cross-disciplinary teams and built working systems across four tracks: Health & Wellbeing, Society & Sustainability, Business & Productivity and Open Innovation. Each track pointed teams at a different kind of problem, but all four came back to a test the summit has used since its earlier events this year: does the technology actually reach the person meant to use it? A patient managing prescriptions, a small business owner reading through customer reviews, a city planning for the day it has to evacuate, all of it counts as the same question.

Six teams made it to the final review, and the panel that scored them looked closer to a product review board than a hackathon judging table. AITEX’s evaluation panel is made up of working professionals, engineers and architects from big tech and enterprise backgrounds who build and maintain systems like the ones these six teams were trying to get right in 48 hours. That background shapes what gets asked. A panel that ships production software all day is asking whether the system underneath a demo would survive a second week of real use.

Every project was scored against five criteria: Innovation and Originality, Real-World Impact and Usefulness, Technical Implementation and Architecture, Feasibility and Scalability, and Presentation and Documentation. Each judge scored each project from 1 to 5 on every criterion, and the final ranking weighted those scores: Real-World Impact and Usefulness counted for 25 percent of the total, Presentation and Documentation for 15 percent, and the remaining three criteria for 20 percent each. It’s a small detail, but it says something about what the summit rewards. Cleverness alone can carry a project partway. Convincing the judges it would matter to someone outside the room is what carries it the rest of the way.

First place went to Bottleneck, an interactive city evacuation simulator, and the margin was not close. The project posted the highest Technical Implementation score of the summit, and judges called the win one of the more clear-cut first-place results the event has seen. Bottleneck is a mesoscopic traffic simulator built for evacuation planning. It models whether a city’s road network can carry the traffic an evacuation would generate, how long full clearance would take, where queues form under pressure, and how a closed road, an added lane or a contraflow arrangement changes the outcome. Roman Postnov, a senior Python developer with more than six years across fintech and startup engineering, built Bottleneck entirely on his own, handling the idea, the build, the testing and the presentation himself. He picked four real cities and set validation targets for each before the event started: San Francisco, Mercer Island, the Florida Keys and Paradise. That choice mattered. It turned the project from a demonstration into something judges could check, comparing the simulator’s output against conditions the team had committed to in advance.

That kind of auditability drew Andrei Privalov, an ML engineer who builds retrieval-augmented generation pipelines and planning agents for a living. He received the summit’s Technical Architecture Excellence Award. Andrei Privalov’s own work involves breaking complex requests into execution graphs a system can reason through step by step, and Bottleneck’s architecture held up under exactly that kind of scrutiny: its assumptions, validation targets and failure points were all sitting in the open rather than buried behind a polished front end. Konstantin Likhonin, who leads transactional products at a bank, has spent his career on systems where failure has immediate, physical consequences, including work that pushed transaction speeds up by a quarter in a high-load environment while holding 99.9 percent uptime. He took the Real-World Impact Award, given for a consistent eye across the summit for the projects most likely to change what happens to real people under real conditions, and an evacuation tool is about as direct a test of that as this field had to offer. Sushrutha Sreevathsa’s Predictive Leadership Award came from a related place. A site reliability engineer turned product strategist who recently finished an MBA, Sreevathsa builds predictive reliability dashboards for a living, which gave him a natural read on a tool whose whole purpose is forecasting how a system behaves under stress that hasn’t happened yet.

Second place went to Platephoto, a food diary built around a photo. Snap a plate, and the app estimates foods, portion sizes and macros, then hands the result to a review screen where the user can correct anything before it saves. Judges liked one design choice in particular: recognition works as an assistant offering a first pass, not a verdict, so a wrong estimate doesn’t just slide through on a task where accuracy matters to someone’s health. Platephoto scored ahead of every project except the winner on both Impact and Technical Implementation. Sambit Panigrahi has spent more than fifteen years leading HR technology transformation in healthcare, including automating benefits administration and payroll inside a heavily regulated environment. He took the Domain Expertise Award for the specialist knowledge he brought to projects like this one, where the stakes sit close to a person’s health but the interface has to stay simple. Konstantin Likhonin’s Real-World Impact lens showed up here too: outside banking, his work has included medical document processing built around measurable gains in accuracy, close to the same test Platephoto’s photo-to-macro pipeline had to pass.

Third place went to Localmind, an offline-first Android field notebook for inspections in places with no connectivity. Record a finding, attach a photo, set a severity and status, search the archive later, all with no account and no network. It posted the strongest Innovation score among the runners-up and finished only four hundredths of a point behind second place, a margin narrow enough to suggest the panel spent real time weighing it against Platephoto. Liudmyla Shumilina, a senior product designer who has scaled SaaS products across more than forty countries and led an accessibility overhaul of an exam-creation flow, took the Innovation Recognition Award. It’s the exact skill Localmind needed recognized: spotting real transformative potential in something that looks, on the surface, like a plain notebook app. Evgenii Poremov, a data analyst specializing in workforce management who has published on scheduling-system architecture, received the Outstanding Evaluation Award for the thoroughness of his written assessments, the kind of close reading a four-hundredths-of-a-point margin tends to require.

Fourth place went to Review Pulse, which takes a pile of dated customer reviews and turns them into recurring themes, supporting evidence and trends, so a small business owner can tell a one-off complaint from an actual pattern. Evgenii Poremov’s own specialty is turning operational noise into metrics and structure a team can act on, which made this project a natural fit for his review: same job, customer feedback instead of workforce data.

Fifth place went to What Changed, which monitors any public URL, keeps clean snapshots of it, and surfaces meaningful changes such as a price shift, a moved deadline or a quietly removed condition. Judges flagged the security thinking built into the crawl path. That’s where Kirill Pechurin came in: a senior backend engineer who built the systems tracking IT assets across more than ten thousand production servers, he received the Responsible Technology Award for how carefully he weighed the ethical and practical dimensions of what he reviewed, a fitting match for a tool whose whole job is watching other people’s pages responsibly.

Sixth place went to Say It Simply, which takes a block of text, a target reading level and a choice of simplify or humanize modes, and rewrites the passage without changing its meaning. Denis Sinyukov leads a software engineering team and mentors developers through architectural reviews as part of his day job. He took the Constructive Guidance Award for the specific, actionable feedback he gave teams across the event, the kind smaller, single-purpose tools like this one often needed most.

Running a six-project final round with a panel this senior takes its own coordination. That job fell to Oleg Bondarchuk, a senior cloud architect with more than sixteen years in hybrid cloud environments who leads a large enterprise’s cloud transformation across more than forty offices. He received the Judging Leadership Award for keeping the panel’s scoring consistent across four very different tracks, so a health tool and a traffic simulator got held to the same bar instead of two different ones. Ratnakar Tripathy, a senior program architect with more than twenty years in enterprise IT and a U.S. patent in intelligent automation that predates the term “agentic AI,” took the Visionary Impact Award for a perspective that shaped how the panel weighed projects well beyond any single track. Mykola Khoroshevskyi has spent his career building distributed systems and data platforms at scale, including earlier senior big data engineering roles before he moved into leading distributed engineering teams himself. That trajectory is exactly what the Systems at Scale Award recognized. Maksim Kananovich has spent more than a decade in C-level roles across the cross-border software companies he’s founded and scaled, guiding distributed teams delivering technical work to international clients, which is the kind of hands-on leadership behind his Mentorship Excellence Award.

Six projects, five criteria, and a panel drawn from people who run production systems at real scale. It’s the same format AITEX has been refining across its recent events, and this time it produced a set of winners whose value is easy to point to, since the scoring was built to reward exactly that. Bottleneck, Platephoto and Localmind carry that recognition into whatever their teams build next, and the rest of the field leaves with a structured, professional-grade read on where their projects stand. The next AITEX Summit will pick up from here. Details on the Summer 2026 event, including the full project list, are available at aitexsummit.com/summer2026.