AI/ML / 2026
TalentAI-X
TalentAI-X was submitted for Tic Tech Toe 26 by a four-member team. We reached the final stage, completed the final presentation, and demonstrated a working multi-agent talent intelligence prototype combining resume processing, skill normalisation, semantic matching, bias-aware evaluation design, and explainable candidate-analysis workflows. The project did not place in the top five.
I am currently preparing a short product walkthrough.
Overview
Candidate evaluation can become overly dependent on keywords and surface-level resume signals, making it difficult to compare evidence from projects, skills, and context. The project structures candidate evidence, normalizes skills, and uses multi-agent workflows to produce clearer fit insights and review artifacts.
My role
- Led a four-member team through planning, build decisions, delivery, and final presentation.
- Contributed across AI workflows, backend APIs, frontend integration, deployment, testing, and demo coordination.
- Helped shape resume processing, skill normalization, semantic matching, bias-aware evaluation design, and explainable candidate insights.
What I built
- Resume processing
- Skill normalisation
- Semantic candidate matching
- Structured candidate evidence
- Bias-aware evaluation design
- Explainability workflows
Project status
Working Prototype. A public demo is available.
What I learned
- AI evaluation systems need evidence, uncertainty, and review paths.
- Talent products require careful wording around fairness, accuracy, and compliance.
Related