I'm a hiring manager. I got tired of reading AI-garbage resumes so I built a tool that does it right.
I built Tamar. It tailors a resume to a specific job description using only your actual experience.
I'm a data science manager at Uber. Every time I open a new role, I get hundreds of applications in the first 12 hours. Not exaggerating, I've watched it happen repeatedly.
At that volume, the first resume screen is a game of probabilities, and those who take an extra effort to match their resume to the actual job description get through that filter more often
But here's what's been driving me insane. I'm not unique in this idea, and recently I've observed a wave of resumes that overfit to the job description in the worst way: people claiming skills they don't have, experiences that never happened, buzzword salads that match the JD perfectly, but not at all - the candidate
When I have 100 resumes to review in a couple hours - this is darn obvious, for me and for any hiring manager.
The frustrating part is that the core idea is right. Matching your real experience to what the role needs? That's genuinely powerful. AI can absolutely help with that. The problem is most tools go straight to fabrication without extensive pre-training and prompting.
So I built Tamar. It build an extensive real user profile, learns what they actually can do, and only than it tailors a resume to a specific job description using only your actual experience. No fake achievements, no exaggeration. It focuses on transferable skills, relevant experience, and the stuff that actually makes you a fit instead of inventing stuff that'll get you caught in the first interview.
It's a side project. Free tier available. Would love feedback from people who've been on either side of the hiring process.