About

The problem

AI agents are being handed real permissions now. They send emails, issue refunds, touch customer records, call outside services. Most teams have no real record of what an agent actually did, and no way to stop a risky action before it happens instead of after.

Tracyn logs every action an agent takes, lets a person step in before the risky ones run, and keeps that record in a form that can’t quietly be edited later. The goal is simple: turn “trust me, the agent is fine” into something you can actually show your team, your customers, or an auditor.

The founder

Tracyn is built by Awais Siddique. I’ve been coding since I was 10. Along the way I taught myself Python, HTML, CSS, JavaScript, TypeScript, and C++, plus frameworks like React, Laravel, and Electron. What started as curiosity turned into shipping full products: AI platforms, desktop apps, browser tools, and machine learning research.

I like problems that are properly hard, things like multi agent AI systems, ICU deterioration detection, and flood response models, and I care about turning them into things people can actually use.

I’m always open to new opportunities, collaborations, and interesting problems to build for.

For more about me and my other products, visit awaissiddique.vercel.app.

Reach me at mawais9171@gmail.com.