TL;DR: We’re testing a new AI-assisted coding tool called Enric Code Assistant that learns the structure, intent, and conventions of your existing codebase, not just the syntax.
If you manage or contribute to a codebase with 30+ developers, we’d love your feedback on how it performs in real-world workflows.
What we’re building
Most AI code tools today focus on autocompletion and snippet prediction.
That’s useful, but it’s not the hardest part of working in large, evolving codebases.
We wanted a tool that could actually reduce the grind of polishing, editing, and making code production-ready.
In order to do that, we figured we needed a tool that is able to read, learn, and adapt code generation to your team’s unwritten rules.
Enric Code Assistant builds a semantic map of your repositories using Abstract Syntax Trees and context fingerprinting.
This allows it to:
- Detect design inconsistencies and code smells specific to your project
- Provide review-level feedback (not just “fix this typo”)
- Reduce onboarding time for new devs by surfacing how things are done here
- Integrate directly with GitHub and GitLab (Bitbucket in progress)
It’s not a chatbot. It’s a quiet, context-aware reviewer that grows with your repo.
Why we’re sharing it here
We’re at the alpha testing stage and looking for teams willing to try it on real-world codebases.
Ideal collaborators are:
- Companies with 30+ active devs
- Teams that already use tools like Copilot, SonarQube, or CodeQL
- Willing to share candid feedback on accuracy, workflow fit, and value
We’re offering free early access for selected teams and direct contact with our engineering crew.
Why this matters
Code generation is easy.
Code understanding is hard.
Our bet is that the next generation of AI tools will be repo-aware, not prompt-aware.
How to join
If this sounds interesting, drop a comment or DM, or fill out this quick form [→ link to Google Form or waitlist].
We’ll review applications weekly and onboard new testers gradually to ensure good support.
Open to discussion
We’d love to hear from you:
- What’s been your main frustration with existing AI coding tools?
- If you could “teach” your assistant one thing about your repo, what would it be?
We’ll stay active in the comments and share early findings as we go.
Top comments (0)