You’ve seen the no-code hype. You’ve seen the AI-in-recruitment posts. Maybe you’ve even dabbled with ChatGPT to write Boolean strings.
But now you want to build something useful, maybe a little tool, a workflow, something that actually helps.
Good. Let’s talk about vibe coding for recruiters
Before we do, I have led about a bajillion tools and systems implementations and here are three rules that I suggest you live by for any TA system, AI enabled or otherwise.
1. You can’t out-tool bad governance.
If your interviewers take a week to give feedback, no AI system will save you. Have the hard conversation before you reach for a tool.
2. Tools ≠ solutions.
Tools make solutions easier to implement. They don’t replace actually having one. Solve the problem first. Then pick the tool.
3. If you can’t draw it, don’t build it.
If you can’t draw your process as a flowchart, then you’re just plugging point solutions in and hoping they work. Learn to map your system. It’s free, and it’ll change how you think.
How to Choose What to Build
If you’re trying to dip your toe into building a tool or workflow, here’s how to narrow it down:
1. Solve a specific, immediate problem.
Example: poor feedback quality, or time-sensitive emails that get missed. Small, scoped problems are buildable. “Fix our entire hiring process” is not specific enough.
2. Look for low logic + high output solutions.
AI tools that use tokens get exponentially more expensive with complexity. So keep the logic simple.
- Low logic: clean up messy spreadsheet data High logic: multi-user analytics dashboards
3. Make it deployable and safe.
Work with your engineering AND legal teams to make sure what you’re building is compliant and secure. DO NOT fly under the radar!
Tools I’d Recommend
For beginners:
- Bolt (my current go-to): clean UI, deployable in clicks, great for web-based tools Lovable: similar to Bolt, beginner-friendly ChatGPT / Claude Code: use these to draft your instructions or help debug
More advanced:
- Cursor & Windsurf + GitHub: proper dev environment, integrates with GitHub, good for long-term or more technical builds
A Few Things I’ve Built (as a recruiter, not an engineer)
- Fee Calculator I built this to show the offset of my budget by offering TA services to eco-system partners. Clean, simple, looks great. Kickoff Co-Pilot Add responsibilities Get success criteria with measurable targets Auto-generate competency questions and RAG-based review guidelines Export as PDF or link → It saves hours, and you can co-create with hiring managers live. Personal Analytics Toolkit (in progress) A personal ATS + productivity suite for individual recruiters: timers, task counters, self-coaching data.
Next goal for me : removing as many SaaS subscriptions as possible and building cheap AI-first replacements. For fun (Miro...I'm coming for you)
Bonus Tip: Use AI to Prompt AI
If you’re using Bolt, don’t wing the prompts.
Grab their docs, paste them into ChatGPT, and ask it to generate a Bolt-friendly prompt tailored to what you want to build. You’ll save tokens, reduce errors, and build faster.
Some Additional Thoughts
Prompt engineering is already on the way out.
Context engineering (prepping docs, diagrams, examples, and schemas) is where we’re headed.
So don’t chase shiny tools.
Map your process.
Solve one problem.
Then build.
It's really fun and it's personally empowering!
Have you been playing with building your own cool wee tools? Tell me all about it! Feel free to DM me on LinkedIn!
Well... that's issue Fifty One! If you have any questions about it, or any feedback on this issue of The Data Driven Recruiter, grab me on LinkedIn for a chat.
I'll see you next week!