CV Management System.
Structured candidate information. Smarter recruitment workflows.
A web-based recruitment management application developed as an academic mini-project, uniting applicant profile management, job postings, skill-based filtering, configurable candidate scoring, and recruiter dashboards into a single workflow.
# 01 // SYSTEM ARCHITECTURE & DATA FLOW
React & TypeScript
Applicant forms · Recruiter hub
TanStack Query
Caching · Mutation invalidation
Weighted Ranking
Skills · Experience · Bonuses
Supabase & Postgres
Auth · Relational schemas · Roles
The project uses Supabase as its primary backend-as-a-service, directly managing PostgreSQL tables for applicants, jobs, and application states. TanStack Query synchronizes server mutations with client-side cache, enabling responsive filtering, real-time status updates, and dynamic candidate sorting without a custom middleware server.
# 02 // DUAL-PERSPECTIVE RECRUITMENT WORKFLOW
Structured multi-step form capturing personal contact information, education history, work experience, technical skills, certifications, references, languages, and career objectives.
Recruiters publish and manage active vacancies with criteria for department, job type, salary brackets, and required skills. Applicants can browse open listings and apply directly.
Recruiters configure scoring weights to evaluate candidates based on role priorities: tuning parameters for skill match, years of experience, certifications, language abilities, and references.
A centralized recruiter dashboard featuring application trends over time, experience-distribution breakdowns, candidate status trackers, and interview coordination tools.
# 03 // CONFIGURABLE CANDIDATE SCORING CRITERIA
Rather than relying on an opaque algorithm, the application implements a transparent, multi-attribute scoring model in SortingPage.tsx. Recruiters adjust slider weights depending on the role's specific needs:
Skills Match Weight
Direct overlap between profile tags and position requirements
Experience Weight
Normalized duration and seniority in relevant industry positions
Role Match Weight
Specific title, department, and candidate objective compatibility
Certification Bonus
Configurable point uplift for verified external credentials
References Bonus
Additional scoring weighting for substantiated manager references
Languages Weight
Evaluation of regional and global multilingual capabilities
# 04 // CHALLENGES & PRACTICAL LEARNINGS
Disparate fields (binary certifications, numeric years of experience, and categorical skill arrays) required mathematical normalization so no single attribute unfairly distorted the ranking curve.
Recruiter updates to candidate stages and interview schedules needed instantaneous propagation. TanStack Query query-invalidation kept dashboard statistics and sorting tables synchronized with zero lag.
Structuring normalized relational tables in Supabase for user profiles, detailed CV records, job postings, application submissions, and interview schedules with relational foreign-key integrity.
Academic Collaboration & Future AI Evolution
Developed collaboratively as a 3-person B.Tech Computer Science and Engineering mini-project submitted to APJ Abdul Kalam Technological University (KTU). My core contributions centered on frontend architecture, Supabase data layer integration, the configurable candidate scoring engine, and the recruiter analytics dashboard.
Because this system was architected before the modern generative AI wave, the next iteration is slated to integrate local LLM agents for automated CV parsing, contextual semantic skill matching (moving beyond strict keyword matching), and dynamic interview question generation tailored to candidate experience gaps.