Some of you may know Greptile started out as a hackathon project in July of 2023 with the simple idea to let developers "talk" to their codebase. 15 months later, Greptile has grown into a codebase intelligence platform - APIs and tools designed to augment modern software teams, centered around AI that understands your entire codebase.
As such, 15 months is not a very long time. Today, we serve over 900 software teams in over 50 countries, from garage-band startups to large enterprises. Our journey of building Greptile has been underlined by nothing if not urgency. We see shipping speed as a core competency, especially as we have been in the stage of quickly validating or invalidating hypotheses around what our customers want. As a consequence, our product felt a little in-cohesive and lacked the level of refinement we wanted to hold ourselves to. With the added load of quickly becoming critical tooling for our customers, it was clear to us it was time to rebuild.
Today, I am excited to show you all Greptile 2.0, a ground-up rewrite of the comprehensive, extensible, and powerful AI-powered platform designed to accelerate your software development life cycle with full-codebase-context at its core.
Greptile's intuitive home interface
Customizable settings for optimal performance
What is Greptile?
Greptile is an AI that understands your entire codebase. It does this by generating a detailed codegraph of how each function, class, file, and directory is connected.
With this full-codebase understanding, Greptile can review pull requests, generate descriptions for tickets, diagnose sentry alerts and failing tests, or integrate with any of your tools via a simple REST API.
How do teams use Greptile?
Software teams use Greptile by embedding it into their workflows. Think of it like an AI staff engineer that understands your codebase, that you can integrate into every step of your SDLC. There are a few ways to do this:
Code Review Bot
Greptile summarizes and reviews PRs to surface hard-to-catch bugs and anti-patterns. Unlike other tools, it has full codebase context, and can identify issues that they miss. Today, we support GitHub and GitLab. Greptile's code reviews can surface:
- Security oversights such as SQL injections.
- Repeated code where an existing abstraction should be used.
- Anti-patterns, such as a new DB connector that is structured differently than the existing ones in the codebase.
- Missed edge cases that may break production (and more!)
Jira/Linear Integration
Greptile can generate descriptions and instructions for JIRA/Linear issues and add them as comments right in the ticket. It does this with full understanding of the codebase, and even leaves links to all the files related to/affected by the ticket.
Custom Integrations
Many teams build custom apps/integrations using our API. Some examples include:
- Connecting Greptile to Sentry or Datadog to diagnose alerts.
- Integrating Greptile with your test pipeline to suggest fixes for failed tests with full codebase context.
- Integrating Greptile with your documentation software to auto-generate wiki articles with codebase context.
Is it private/secure?
Greptile is SOC2 Type II compliant and undergoes annual external audits and rigorous penetration testing. Our cloud product does not store customer code, instead pulling snippets as needed from your code host, and Greptile can also be self-hosted in customers' own cloud as part of our enterprise plan.
To try greptile, head to greptile.com. To talk to someone from our team, email us at hello@greptile.com.
Week 1, 2am at the office, quick break to listen to Charli XCX's boiler room set.
Week 4, 3:30am at the office, barely alive.