Vstorm Academy

Everything we learned shipping agents to production, written down.

The blog, the glossary, the guides we hand clients during delivery, and the agent libraries we maintain in public. Most of it needs no email address.

Read the blog Browse the glossary
Start here

Six ways in

Filter by format, sort by size, and peek inside each library before you open it.

Not sure where to start? Talk it through with an engineer

Showing 6 of 6

Watch

Two clients, on camera

Both were recorded with the client after the system was running in production. Neither is a demo reel or a scripted testimonial.

  • Synera

    Building text-to-workflow Agentic AI with Synera

    Engineers moved from hours of tedious setup to minutes, through multi-step validation rather than a single generation pass.

    2 hrs → 3 min to generate a validated workflow
    Read the case study
  • Mixam

    Multi-agent AI-support facilitating highly customized order completion for Mixam

    A three-agent product advisor guiding customers through print-order configuration — 15 tools working against more than a billion product combinations.

    95.4% Success rate in workflow results
    Read the case study
More on YouTube
Tools

Two things you can run yourself

Start with a guided readiness profile, or jump straight into the agent libraries we ship in public.

Interactive

AI readiness assessment

Profile business, data, technology and governance readiness in one guided pass — then leave with a structured score.

  • 10–15 min
  • Free
  • No scoping call
Open assessment
Open source

Agent tooling on GitHub

30 public repositories and 3,700+ GitHub stars, built on Pydantic AI — including Pydantic DeepAgents. Counted on github.com/vstorm-co · 26 August 2026.

  • 30 repos
  • 3,700+ stars
Browse on GitHub
From the blog

Most recent articles

The blog is the most active surface here — these three are the newest.

All articles
RAG

Is higher accuracy always the goal?

A “minimum 95% accuracy” promise can be worth nothing. An interactive walkthrough of precision, recall and what to ask before any number goes in writing.
Agentic AI

AI agents can act: can you still control them?

Agents that issue refunds need identity, authorisation, approval and audit before the call lands. What an agentic AI control plane is, and when you need one.
Agentic AI

Insurance fraud detection with AI: what works and where it fits

How insurance fraud detection works with AI and agentic systems: text analysis, cross-source correlation, identity screening and real-time claim intake.
What is here

A library, not a landing page

Almost all of it is free to read, no email required. The counts come from the content itself, so they cannot drift from what is published.

241
Articles on agentic AI delivery, architecture and evaluation
686
Glossary entries defining the agentic AI vocabulary
30
Open-source repositories our engineers maintain in public
Newsletter

Two agentic-AI dispatches a month

Every two weeks on Wednesday: plain-English breakdowns of what we ship, from CEO and founder Antoni Kozelski and the delivery team. No spam, unsubscribe anytime.

FAQ

Questions about the Academy

What is in Vstorm Academy? +
Four things: 241 blog articles on agentic AI delivery, an AI glossary with 686 entries, long-form guides and ebooks, and 30 open-source repositories our engineers maintain. It also collects the client story videos and the AI readiness assessment.
Do I have to give you my email? +
Not for most of it. The blog, the glossary, the open-source repositories and the videos are open. The ebooks and the RAG guide are gated behind an email form, because they are the lead magnets that fund the rest.
Is this a course with lessons and a certificate? +
No. Academy is a library, not a curriculum. There is no enrollment, no sequence, and no certificate. It is the written record of what we learned building agentic AI systems for clients, published so you can read it without booking a call.
Who writes this material? +
It depends on the piece, and the byline says which. Our founders and research analysts write most of the strategy and market analysis, the content team writes the explainers and the glossary, and the engineers who build the systems write the deep technical posts — those also run on oss.vstorm.co. Case study numbers come from the projects themselves, and client quotes are published with a name and a title.
How often does it change? +
The blog is the most active surface. The glossary grows as terms come up in delivery. Open-source repository counts and star totals are point-in-time facts, refreshed when we recount them — the current figures were counted in July 2026.
Can we get a workshop instead of reading? +
Yes. The Pragmatics of AI workshop covers the same ground in a session run with your team, and the AI readiness assessment gives you a structured profile in 10 to 15 minutes. Both are linked from this page.
Work with us

Read enough? Bring us the workflow that is costing you the most

A 45-minute call with an engineer who assesses whether an agent fits your workflow.

Book a discovery callSee case studies