08:00
→ 09:00
Check-in & Welcome Coffee
Doors open at 08:00. Volunteers scan tickets at the entrance. Coffee, tea and pastries in the networking area — use the hour to meet the first attendees.
Program
The agenda runs on two parallel stages, Wednesday 24 and Thursday 25 September, from 09:00 to 18:00. Keynotes and panels on the main stage; 30-minute technical sessions on the second stage. The final agenda is published one month before the event.
08:00
→ 09:00
Doors open at 08:00. Volunteers scan tickets at the entrance. Coffee, tea and pastries in the networking area — use the hour to meet the first attendees.
09:00
→ 09:30
Official welcome, house rules and what to expect from the two days. We announce the tracks, lead sponsors and how curated networking works.
09:30
→ 10:15
An honest look at how AI is reshaping business models: what changed in the last 12 months, what will change in the next 12, and how to position your company so you don't have to rebuild it from scratch.
10:15
→ 11:00
GPUs, vector DBs, retrieval pipelines, observability. A technical tour of the infrastructure stack that powers modern AI products — with real numbers on cost, latency and throughput.
11:00
→ 11:30
Coffee, fruit, savoury snacks. Volunteers facilitate 1:1 introductions based on the preferences you declared at check-in.
11:30
→ 12:15
Three companies that took AI out of the POC stage and made it deliver business results: how they structured the team, what metrics drove their go/no-go calls, what they cut to ship on time.
12:15
→ 13:00
Unvarnished conversation with a founder who built an AI company: the costly mistakes, the moments they were tempted to quit, and what they would do differently starting today.
13:00
→ 14:00
Catering included with every ticket. Tables of 8 grouped by topic (infra, sales, talent, M&A) — pick where to sit.
14:00
→ 14:45
A concrete framework for the CEO who must decide: buy, build, or partner? How to prioritize high-impact use cases, measure ROI, and which capabilities to bring in-house.
14:45
→ 15:30
Technical session in the Deep Dive Room: architecture of a production agent — tool use, planning, eval, guardrails. Real code, not marketing slides.
15:30
→ 16:00
Coffee, sweets, sparkling water. Partner booths open in the networking area.
16:00
→ 16:45
A senior VC on current theses: which AI companies attract capital, which moats are real and which are illusions, current valuation multiples, and where they see opportunities in Central Europe.
16:45
→ 17:30
How to align an AI system with your company's values, what logs to keep, how to respond to incidents. Honest discussion of what works in practice, not just on paper.
17:30
→ 18:00
Recap of Day 1 key ideas, Day 2 announcements and after-party logistics. Head to your hotel or the old town for dinner.
08:00
→ 09:00
Same logistics as Day 1. If you missed a session yesterday, volunteers can give you the quick highlights.
09:00
→ 09:30
Five-minute recap plus a Day 2 overview — which sessions to prioritize if you're a CTO, a PM, or a founder.
09:30
→ 10:15
How to recruit, retain and grow engineers in a market where every big tech is chasing the same 200 people. What works, what doesn't, and why "remote vs office" is no longer the question that matters.
10:15
→ 11:00
A case study from a 1,000+ employee company that integrated AI across every department: what worked, what blew up in their face, and how it shows up in the P&L a year later.
11:00
→ 11:30
Strategic break — we use it to introduce you to the morning's speakers.
11:30
→ 12:15
Three highly regulated industries, three different approaches to AI. What's allowed, what's forbidden, and how to deliver real value within those limits.
12:15
→ 13:00
EU AI Act, GDPR, NIS2, DSA — what they mean concretely for a tech company in Romania. With a lawyer working on these files day in and day out.
13:00
→ 14:00
Generous buffet, vegan and gluten-free options. If you have special requirements and didn't tell us, stop by the info desk — we'll sort it out.
14:00
→ 14:45
Llama, Mistral, Qwen vs GPT, Claude, Gemini. When to go self-hosted, when to use the API. Cost, latency, quality — and what happens to your IP in each case.
14:45
→ 15:30
How to cut inference cost 40-70% without losing quality: prompt compression, model routing, aggressive caching, batching, targeted fine-tuning.
15:30
→ 16:00
Final break of the event — use it for the contact exchanges that matter.
16:00
→ 16:45
Prompt injection, data leakage, supply chain attacks on models and plugins. With real examples and a hardening checklist you can apply immediately.
16:45
→ 17:30
Four Romanian founders building AI products for the global market. How to attract international customers from Bucharest, what to hire locally and what you need from abroad.
17:30
→ 18:00
The last five ideas to take home, thanks to speakers and volunteers, and the official invitation to the after-party.
19:00
→ 23:59
Final location announced on stage at closing and via email — a bar 10 minutes' walk from Hala Laminor. Open bar until 22:00, half-price drinks after with your badge. Unstructured networking, no agenda.
Working program