Training

What we train, and how.

No lecture theatre. Trainees learn AI by shipping real commercial slices under senior supervision — the same curriculum whether you join the accelerator or hire a graduate.

Apply to the accelerator

Six tracks.

Every trainee gets grounding in all six, then specialises through the projects they're assigned to.

Prompt & agent design

Reliable prompting patterns, tool-use, memory, and goal-driven agents you can actually ship.

Workflow automation

n8n, Zapier and Make — plus custom scripts — to replace repetitive work across Microsoft 365 and Google Workspace.

Data plumbing & APIs

Clean integrations, webhooks, enrichment and reporting between the tools clients already use.

RAG & knowledge assistants

Index, retrieve and ground answers over Drive, Notion and internal docs — surfaced in Slack, Teams or the web.

Evals & guardrails

Write eval suites, red-team, and add safety rails so what ships keeps working after handover.

Client craft

Discovery calls, scoping, written briefs, demos and handover — the professional skills that turn engineers into deliverers.

How the training runs.

Foundations week

Intensive grounding in Python, cloud, LLM APIs, git and the tooling every squad uses. Same starting line for everyone.

Live project pods

Small pods pair on internal builds first, then real client slices — reviewed weekly by a senior AI engineer.

Client-facing delivery

You run stand-ups, write the brief, demo the work. Supervised, but visibly leading — the reference every hiring partner wants.

Tools trainees leave fluent in.

Placeholder stack — swap for real partner logos as commercial engagements land.

OpenAIAnthropicn8nSupabaseVercelZapierHubSpotAirtable

Deep-dive modules

Companion training on aiacceleratorkit.com

We keep an open library of role- and topic-specific modules on our companion kit. Trainees use them; clients are welcome to as well.

Companion training hub

The full training catalogue on our companion kit.