From first prompt to multi-agent systems.
Some teams need everyone set up on Claude and using it well by Friday. Others already run skills and evals and want multi-agent loops, RAG, deeper automation. Every program is composed for your people and the work they already do. Led by an ex-Microsoft GenAI lead with several patents in multi-agent systems.
No two programs are the same.
We compose each program from three things: how many people you have, what they already do with AI, and what you want out of it. Some teams start at setting up Claude. Some have been running assistants for a year and want agent architecture.
Every session is hands-on. Participants work on their own tasks, in their own tools, so what they learn on Tuesday is in use by Thursday. Delivered onsite or remote, over a few weeks.
What we offer.
Programs pull from these, in whatever order fits your team.
-
Team onboarding
Everyone set up on Claude and Codex, using them properly and safely in daily work.
-
Bring your own workflows
Each participant brings a real recurring task and leaves with it about 30% faster.
-
Quality control and evals
Fact-checking, review checklists, and evals in plain English: knowing when an output is good enough.
-
Multi-agent systems
Agent loops, orchestration, and handoffs. When a system of agents beats a single model.
-
Retrieval
RAG and graph RAG: getting your firm’s knowledge into the model reliably.
-
Workflow library
Your team’s strongest workflows turned into shared assets: skills, templates, prompts, checklists.
-
Executive sessions
Where AI fits your organization, what to automate, and what stays human-reviewed.
-
Automation roadmap
A leadership view of what to scale next and what should stay with people.
The workshops stand on their own.
Most teams stop there, better at their jobs. Some keep going: the strongest workflows become a shared library, the library becomes organizational memory, and the memory becomes agents that run the work end to end. When a team wants that layer, we build it.
The details, up front.
How is the program composed?
On a scoping call. We look at how many people you have, what they already do with AI, and what you want out of the program, then propose the workshops and the order. You have the full picture, in writing, before anything is booked.
Remote or onsite?
Both. Onsite works best for the hands-on sessions; remote works fine for most teams. Programs typically run over a few weeks so people can use what they learned between sessions.
What team sizes do you work with?
From a handful of people to the whole company. Larger groups split into cohorts so every participant still works on their own tasks.
Which tools do you teach?
Mostly Claude and Codex, plus whatever your team already uses alongside them. We are model-agnostic and take no referral fees from any provider.
What do participants need to prepare?
Nothing beyond a short intake. Each person arrives with one real task they want to improve, and we take it from there.
What does it cost?
It depends on the composition: how many workshops, how many people, onsite or remote. Exact pricing goes in the proposal after the scoping call.
We’re in a regulated industry. Does this still work?
Yes. The program is built around your compliance constraints: what can and can’t go to a model gets settled up front, and the quality-control sessions exist for exactly this reason.
What happens after the program?
Whatever you want it to. Most teams keep running on their own. Some turn their best workflows into shared firm assets, and some go further into agents we build. The workshops stand on their own either way.
Start with a scoping call.
Thirty minutes. You tell us where your team is; we tell you what we’d compose and why. If workshops aren’t the right move for your team yet, we’ll say so.