AI Candidate Sourcing Automation

via Freelancer ·

Budget / Salary₹30,000–250,000
TypeFreelance project
LocationRemote
Posted42 minutes ago
The objective is to automate as much of our recruitment process as possible to improve
recruiter productivity, reduce manual effort, speed up sourcing and bring greater efficiency to
the overall recruitment cycle.
We aim to build an AI-powered recruitment platform that integrates multiple job portals with
our in-house ATS, and automates recruitment tracking and reporting.

A. Centralised AI-Powered Job Search
Develop a new search interface with Claude AI integrated at the backend, connecting to the
following job portals:
 Naukri
 Shine
 Indeed
 LinkedIn
The system should:
1. Allow recruiters to search across multiple job portals from a single interface.
2. Accept Job Descriptions (JDs), keywords, skills, experience, location and other
relevant search criteria.
3. Use AI to understand the JD and identify relevant candidate profiles.
4. Display profiles ranked by their relevance to the requirement.
5. Eliminate the need to individually visit and search each job portal.

B. Candidate Profile Display & Shortlisting
The search interface should display candidate profiles in a list/summary format rather than
automatically opening individual profiles.
 Mobile numbers and contact details should not be visible in the search results.
 Profiles should not open automatically.
 Recruiters should be able to review basic candidate details and AI match scores.
 Recruiters will manually open only the profiles they wish to shortlist.
 Shortlisted profiles should then be added to our ats system recuitZ.
Integration nd automation
Automated MIS and weekly reporting

Core scope
• Build or train an AI/ML routine (Python, Node, or comparable stack) that scans job boards, LinkedIn-style networks, GitHub, and other open sources to identify profiles matching a role description I feed it.
• Extract key fields—name, headline, skills, location, email or alternative contact, and profile URL—and normalise them to Recruit’s data model.
• Push the enriched records into Recruit in real time, handling duplicates gracefully.
• Provide a simple dashboard or command-line report that shows the number of leads pulled, success/error logs, and API response codes for traceability.

Deliverables
Phase 1 AI-powered search interface and job portal integration
Phase 2 Candidate shortlisting and SpotRecruit integration
Phase 3 Automated MIS, dashboards and weekly reporting
Phase 4 Process optimisation and additional AI-based automation

Acceptance criteria
• At least 80 % of captured profiles populate correctly in Recruit with no manual cleanup.
• Duplicate detection works against existing candidate IDs or email addresses.
• Runtime for a single search stays under 5 minutes for 1,000 profiles.

When you respond, highlight relevant experience building similar sourcing bots, Recruit (or comparable ATS) integrations, and any AI/NLP models you have deployed in production.
website design
Apply on Freelancer →

Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.