The Problem That Hit Home
This project came from my own frustrating experience job hunting. I went through over 80 interviews and 200+ HR calls, but hardly anything worked out. The whole process felt broken - applications disappeared, feedback was generic, and most opportunities were terrible matches.
Then I realized something simple: we use chat for everything now. Google searches, asking questions, getting help. So why was job searching still stuck with endless forms and keyword matching? Why couldn't I just chat about what I wanted and find the right job? Why couldn't recruiters describe their ideal candidate in normal language and get matches?
That's how Chatcruit started - from wanting job searching to be as easy as having a conversation.
What Chatcruit Actually Does
Chatcruit is an AI recruitment platform that works like having a smart career assistant.
For job seekers: Upload your resume, certificates, even old salary slips - the AI reads everything and understands your skills. Then just chat to find jobs, get resume tips, or practice interviews. No more tweaking resumes for every application.
For employers: Describe your ideal candidate in plain English. The AI finds the best matches from real job seekers who are actually looking. No more sorting through hundreds of random applications.
It's about replacing the frustrating application process with simple conversations.
How I Built It (With AI Help!)
Here's the interesting part - I used AI to build an AI project. This meant a lot of back-and-forth with AI agents, debugging their suggestions, and dealing with all the usual AI quirks while trying to build something that uses AI well.
The tech stack:
- Next.js 14 and TypeScript for the foundation
- Tailwind CSS and shadcn/ui for the interface
- Supabase for the database, authentication, and file storage
- pending for the chat and document analysis
- Bolt.new as my main development platform
Bolt.new was a game-changer here. As a solo developer, it let me focus on the complex AI stuff instead of getting stuck on basic setup. But even with these tools, building an AI product using AI tools created some interesting challenges.
The Real Challenges
Building this solo with AI assistance brought unique problems:
AI Building AI Problems: Getting AI tools to help build AI features was like inception - lots of circular conversations and debugging AI-generated code that was supposed to work with other AI systems.
Document Chaos: Making the AI understand different file types (resumes, certificates, pay stubs) while keeping everything private was harder than expected.
Making AI Feel Natural: Building chat that feels like talking to a person, not a robot, took many iterations and prompt adjustments.
Two Different Users: Job seekers and employers need completely different experiences on the same platform.
Solo Development Reality: Managing a complex project alone meant wearing every hat - developer, designer, tester, and product manager.
The irony wasn't lost on me: I was constantly going back and forth with AI agents to solve problems, while building a platform designed to eliminate back-and-forth problems in recruiting.
What I'm Proud Of
Solving My Own Problem: Turning personal frustration into something that could help others avoid the same pain.
Building Complex AI Solo: Using AI tools to build an AI platform as one person feels like a small miracle of modern development.
Making AI Simple: Despite all the complex tech underneath, the user experience is just chatting normally.
Privacy First: Built with strong privacy controls from the start, not as an afterthought.
What I Learned
AI Development is Meta: Using AI to build AI products creates unique challenges you don't expect.
Simple UI, Complex Backend: The best AI experiences feel effortless to users, even when there's sophisticated tech running behind the scenes.
Iteration is Everything: Especially with AI features, you have to test, adjust, and test again constantly.
Modern Tools Enable Big Dreams: With platforms like Bolt.new and cloud services, one motivated person can build enterprise-level software.
What's Next
This is just the beginning. I want to add:
- Smarter AI: Better career path predictions and personalized skill recommendations
- Mobile Apps: Making this available everywhere people actually look for jobs
- Enterprise Features: APIs and integrations for bigger companies
- Community: A place where people can share job search experiences and help each other
The goal is simple: make finding the right job or candidate as easy as having a conversation.
Built With
- ai
- bolt
- claude
- nextjs
- react
- supabase
- tailwindcss
- typescript
Log in or sign up for Devpost to join the conversation.