Harness Institute trains teams to build software with AI.
Hands-on workshops with Claude Code, Cursor, Codex, Devin, Gemini and GitHub Copilot. Built around your team’s own work and tools. See the workshops.
Founded by ambassadors for OpenAI, Anthropic, SpaceXAI and n8n, we train engineering, product and design teams. We help teams move from individual experiments to a shared way of working with AI.
Trusted by
Common blockers after a team adopts AI
A few people move faster. The team does not.
Writing code is faster. Reviewing it takes longer.
The demo works. Your codebase is different.
The feature is built. The requirements are still unclear.
Your team is unsure what AI is allowed to do.
You pay for the tools. The return is hard to show.
Do not let another sprint go to trial and error. Bring the task that keeps getting stuck. We will help you find where to start. Book a call, or measure where your team stands with the team proficiency matrix.
Measure where your team stands
Most teams we meet have the tools but no shared way of working with them. Some engineers hand whole tasks to agents. Others still use AI to autocomplete a line. Before we plan a training, we measure where each person stands.
The team proficiency matrix does that in five levels. Each engineer answers a short questionnaire about how they work today. You get a team report with your current level and the steps to the next one. We use the same report to shape the training around your team.
How we work
An engagement follows six steps: a 30-minute introductory call, the team proficiency matrix, a stakeholder call, a written proposal, acceptance and scheduling, then bespoke training based on your codebase. The formats we deliver are on the training page.
- Introductory callA 30-minute call to discuss objectives, stakeholders and the work to be improved. Book the call.
- Team proficiency matrixEach team member completes a short assessment, so the programme matches the team's actual level.
- Stakeholder callA call with the owners of requirements and review, to align perspectives and priorities across the business.
- Written proposalA proposal with scope, schedule, materials and total price, for your review before acceptance.
- Acceptance and schedulingDates and preparation are confirmed, including the access and materials your team needs to arrange.
- Bespoke trainingTraining based on your codebase, onsite, virtual or offsite.
AI Coding Workflow for Teams
11.09.26
Measure AI Coding ROI on Teams
11.04.26
Choose a Code Review Model
10.31.26
Compare AI Coding Tools by Workflow
10.31.26
AI coding training for engineering, product and design
10.30.26
Data boundaries and human approvals for AI coding agents
10.30.26
From AI experiments to a team workflow
10.30.26
How Autonomous Coding Agents Write Code
10.30.26
How to review AI-generated code
10.30.26
ROI of AI coding tools
10.30.26
Shared standards for Claude Code, Cursor and Codex
10.30.26
Skill Rubrics for AI Coding Teams
10.29.26
Choose an AI Code Review Workflow
10.12.26
AI Code Assistant Training Plan
10.11.26