AI-Driven PHP ATS Webpage
Budget: $10 – $40 USD
I’m building a PHP-based webpage that behaves like a full-fledged Applicant Tracking System. The core of the project is an AI layer that can 1) parse and analyse incoming résumés, 2) score and rank every candidate strictly on skills match, and 3) push approved vacancies live by reading directly from our internal job database.
Here’s the flow I have in mind:
• Recruiters drop a PDF or DOCX résumé onto the page; the parser immediately extracts skills, experience dates and keywords, then writes the structured data to MySQL.
• A ranking routine compares those extracted skills against each role’s predefined skill set, weights the relevance, and returns an ordered shortlist with an explain-score for transparency.
• When a new role is added inside our existing internal database, the same PHP page should automatically create the posting, complete with metadata, without anyone duplicating effort elsewhere.
All server logic must stay in PHP (Laravel or plain PHP with Composer libraries are both fine) but feel free to spin up a lightweight Python micro-service or call an external NLP API if that accelerates résumé parsing accuracy. I already host on an Apache/LAMP stack, so deployment should be straightforward.
Deliverables (acceptance criteria):
1. Clean, commented PHP source code plus any auxiliary scripts.
2. MySQL schema and seed data for roles, skills and candidates.
3. Working résumé parser that achieves at least 90 % field-extraction accuracy on a mixed test set.
4. Skill-based ranking algorithm returning a JSON-ordered list with score explanations.
5. Automated job-posting module that reads from the internal DB and publishes within the same interface.
6. Basic, responsive UI so my recruiters can test everything end-to-end.
If you’ve integrated AI or NLP into PHP projects before and can hit the accuracy benchmark above, let’s chat about timelines and milestones.
Here’s the flow I have in mind:
• Recruiters drop a PDF or DOCX résumé onto the page; the parser immediately extracts skills, experience dates and keywords, then writes the structured data to MySQL.
• A ranking routine compares those extracted skills against each role’s predefined skill set, weights the relevance, and returns an ordered shortlist with an explain-score for transparency.
• When a new role is added inside our existing internal database, the same PHP page should automatically create the posting, complete with metadata, without anyone duplicating effort elsewhere.
All server logic must stay in PHP (Laravel or plain PHP with Composer libraries are both fine) but feel free to spin up a lightweight Python micro-service or call an external NLP API if that accelerates résumé parsing accuracy. I already host on an Apache/LAMP stack, so deployment should be straightforward.
Deliverables (acceptance criteria):
1. Clean, commented PHP source code plus any auxiliary scripts.
2. MySQL schema and seed data for roles, skills and candidates.
3. Working résumé parser that achieves at least 90 % field-extraction accuracy on a mixed test set.
4. Skill-based ranking algorithm returning a JSON-ordered list with score explanations.
5. Automated job-posting module that reads from the internal DB and publishes within the same interface.
6. Basic, responsive UI so my recruiters can test everything end-to-end.
If you’ve integrated AI or NLP into PHP projects before and can hit the accuracy benchmark above, let’s chat about timelines and milestones.