TalentAIQ Lite is live: local-first AI coding evidence for technical interviews

[Lling0000]

TalentAIQ Lite is now live.

TalentAIQ Lite is a local-first, candidate-authorized, redaction-by-default evidence generator for AI coding interviews. It reads candidate-authorized local Codex records, Claude Code records, local Git data, and local project traces, then produces an interview-ready evidence packet.

What it generates:

• Markdown evidence report
• Structured JSON
• Interview follow-up questions
• Privacy checklist
• Poster HTML
• SVG interview profile

What it does not do:

• No automatic candidate ranking
• No automatic rejection
• No hiring recommendation

Why I built it:
AI coding is now part of real engineering work, but "used AI a lot" is not the same as "can ship reliable software". Interviewers still need inspectable evidence and clear follow-up questions.

Current highlights:

• Local-first by default
• Candidate authorization required
• Redaction by default
• Supports Codex, Claude Code, local Git, and project traces
• README now includes a real output preview
• AI JSONL collection now defaults to full scan without hard caps

Repo: https://github.com/Lling0000/TalentAIQ

If you work on AI-native hiring, technical interviewing, developer tooling, or privacy-first workflow design, I’d love feedback.