WealthTech · Wealth-management software
Industrialising agentic AI across engineering.
A shared agentic-AI practice, deployed across more than 20 engineering teams, from AI-assisted coding to vulnerability remediation, with common standards.
Rolled out to pilot teams, then at scale
Standardised internal AI practices
Reusable foundation from one use case to the next
Context
Harvest is the French leader in wealth-management software: its solutions equip wealth advisers, family offices, private banks and insurers, with more than 4,600 client companies and over 600 employees. At the heart of its strategy: industrialising AI to serve engineering, security and software quality, well beyond a single pilot team.
The challenge
- AI adoption still confined to a few pilot teams
- No common standards for usage, security or governance
- A priority security challenge: spotting and remediating vulnerabilities faster
What we delivered
- “AI for Engineering” standards: AI coding assistants, shared conventions and governance
- Specialised agents: code review, tests, documentation, vulnerability remediation
- Agentic platforms on Amazon Bedrock, integrated into CI/CD pipelines
- Adoption supported across more than 20 teams: training, coaching, usage tracking
The result
Teams rely on a shared foundation: each new use case plugs in faster than the last.
Why it matters: AI doesn't stay confined to a pilot team; it becomes a shared, governed and repeatable practice.
What it proves: Intelligence Partners can spread a consistent, governed AI practice well beyond a single team.
Technical detail
Amazon Bedrock · GitLab Duo · AI coding assistants · MCP connectors · CI/CD pipelines · Secure access
Mission led by
Dorian Richard
Founder · Intelligence Partners