Sanwi - Software documentation with built-in app security testing

Hi PH 👋

I’m Landry, CEO & co-founder of @Sanwi.

We didn’t set out to build another dev tool or security scanner.

At first, we just wanted to make software documentation easier for our own team. But after talking to more teams, it became clear the real problem wasn’t docs. It was understanding.

Code ships faster than ever (especially now with AI). Docs fall behind. Security tools flood you with alerts. And over time, no one really knows how the system fits together or what actually matters.

The problem is clear:
Outdated documentation → poor knowledge transfer → missed security risks → slower shipping.

Existing tools treat documentation and security as separate silos. You get doc generators that don't understand your architecture, and vulnerability scanners that flood you with 500 "critical" findings when only 12 are actually exploitable.

That's why we built @Sanwi

@Sanwi is an AI platform that understands your codebase and keeps everything in sync: docs, architecture, and security.

You connect your repos, and @Sanwi does the rest:

🧠 Our AI agents actually understand your codebase not just syntax, but the why behind it. Business logic, architecture decisions, and how everything connects.

📝 Living documentation (Technical architecture document, Developer and User guides, API references etc.) auto-generated and always in sync with your codebase. Google Docs-style real-time collaboration for your entire team.

🔒 We Combine SAST + SCA scanning enriched with CISA KEV and EPSS exploitability data. Our AI agents translate scan results into plain-language explanations with concrete fix guidance, prioritized by real-world risk (not just CVSS scores). Because the same AI agents that understand your code also understand your vulnerabilities; so you can see:

• where a vulnerability lives

• how it connects to real data flows

• whether it’s actually exploitable

• and how to fix it in context

Instead of noise, you get a prioritized, developer-friendly path to what actually matters.

One thing we’ve learned building this:
the problem isn’t a lack of tools rather it’s that they don’t actually understand the system they’re analyzing.

We think the future isn’t more alerts, it’s context-aware.

@Sanwi is our take on that.

We’re just getting started, so would genuinely love your feedback 🙏

Drop your questions below... we'll be here all day! 🚀