SkillGap Radar

Inspiration

Hiring today is still broken. Most Applicant Tracking Systems rely on keyword matching, which fails to capture real capability. A candidate may understand container orchestration deeply but never explicitly write “Kubernetes” on their resume — and gets filtered out.

As a student engineer preparing for industry roles, I experienced this gap firsthand. I wanted to build a system that evaluates candidates the way a senior hiring manager would: by understanding context, depth, and risk — not keywords.

That led to SkillGap Radar.

What it does

SkillGap Radar is an AI-powered decision-support system that performs semantic skill gap analysis between a Job Description and a Candidate Profile.

Instead of asking “Do these words match?”, it asks:

What skills are implicitly required?
What level of competency is expected vs. demonstrated?
What business risk does a missing skill introduce?
Can the gap be closed with targeted upskilling?

The output is a clear executive dashboard showing match score, critical gaps, risk areas, and a personalized learning pathway.

How we built it

The application is built as a modern single-page web app using React + TypeScript and powered by Google Gemini 3.

At the core is a carefully engineered Gemini prompt that:

Infers implicit skills from the JD
Scores required vs. observed competency on a 1–5 scale
Calculates gaps mathematically
Explains reasoning with evidence
Generates an actionable learning plan

The Gemini API is used in structured JSON mode, ensuring reliable, schema-safe outputs for visualization. Optional “Thinking Mode” allocates a reasoning budget so the model deeply analyzes profiles before responding.

Results are visualized using radar charts, gap matrices, and risk indicators to make decisions immediately actionable.

Challenges we faced

Prompt reliability: Ensuring Gemini always returns valid structured JSON required strict schemas and defensive prompt design.
Resume parsing: Supporting PDF, DOCX, and TXT formats while preserving semantic meaning.
UX clarity: Presenting complex AI reasoning in a way that non-technical hiring managers can understand in seconds.

What we learned

Semantic reasoning dramatically outperforms keyword matching for hiring decisions.
Structured AI outputs are essential for production-grade applications.
Gemini 3’s long-context reasoning enables entirely new classes of decision-support tools.