AI-Powered Recruitment ATS Platform by Abdul MannanAI-Powered Recruitment ATS Platform by Abdul Mannan

AI-Powered Recruitment ATS Platform

Abdul Mannan

Abdul Mannan

Overview

Built a full-featured Applicant Tracking System (ATS) for a staffing firm managing 50+ hiring clients. The platform replaced a fragmented spreadsheet-and-email workflow with a unified, AI-driven hiring pipeline.

The Problem

The client was managing hundreds of job applications per week across email inboxes and shared spreadsheets. Duplicate candidates, missed follow-ups, and no reporting visibility were costing them deals.

What I Built

AI Resume Parser — integrated OpenAI API to extract structured candidate data (skills, experience, education) from uploaded PDFs and Word docs with 90%+ accuracy
Smart Candidate Ranking — vector similarity scoring matched candidates to job requirements automatically, surfacing the top fits without manual review
Multi-stage Pipeline — configurable Kanban-style pipeline (Applied → Screened → Interviewed → Offered → Hired) with automated status notifications via SendGrid
Client Portal — white-labeled portal for each hiring client to review shortlisted candidates, leave feedback, and approve offers
Analytics Dashboard — time-to-hire, source tracking, pipeline conversion rates, and recruiter performance metrics

Tech Stack

Frontend: React, TypeScript, Tailwind CSS, React Query Backend: .NET Core 8 Web API, Entity Framework Core AI: OpenAI API (GPT-4o for parsing, text-embedding-3 for ranking), ABBYY OCR Database: PostgreSQL, Redis (queue & cache) Infrastructure: AWS (EC2, S3, SES, Lambda), Docker, GitHub Actions CI/CD

Outcome

Reduced average time-to-shortlist from 3 days to under 4 hours. The client onboarded 12 new hiring companies within 2 months of launch, citing the platform as a direct sales differentiator.
Like this project

Posted May 14, 2026

AI-powered ATS for staffing firms: resume parsing, candidate ranking, and automated shortlisting.