The $500K deck trap
Why most mid-market AI projects die before shipping — and the structural reason it keeps happening.
Agentic AI, shipped — not presented. I run end-to-end agentic AI transformations for mid-market companies. I write and teach what I learn from the work. On the side, I back B2B agentic AI founders.
A small set of mid-market transformations, delivered end-to-end with the team I lead at Vstorm. Outcomes in production — not slideware.
"Antoni's team turned a six-month roadmap into a six-week production rollout. Outcome-first, no theatre."SM Sara Meyer VP Ops · SaaS · 1,400 FTE
"They ship. We had a working agent in our customer ops stack before most vendors had finished their discovery deck."JT Jonas Thiel COO · Manufacturing · 2,200 FTE
"Strategic partner, not a contractor. The work paid for itself inside the first quarter post-handover."RA Rania Al-Khatib CEO · Fintech · 600 FTE
I partner with a small number of mid-market companies each year on end-to-end agentic AI transformation. One team, one operator, one outcome — delivered by the team I lead at Vstorm.
I'm inspired by small giants. I'm building one, and agentic AI is the best tool I've ever had for helping other companies do the same.
Read the full story →Small checks, fast decisions, and real operator help beyond the capital. I invest in one field — the field I work in every day.
See thesis + portfolio →Why most mid-market AI projects die before shipping — and the structural reason it keeps happening.
The boutique cadence we use for mid-market transformations — and where the time actually goes.
A decision tree for 200–5,000 person companies sitting on the agentic AI fence.
One short, useful note every other week — what's shipping, what's breaking, and what's actually working in agentic AI for mid-market.
Long-form conversations with mid-market operators shipping AI in production. Plus the occasional live build.