Overview

The UK-based leading healthcare staffing and recruitment organization was looking to process a rapidly growing volume of resumes, job postings, and candidate-to-job matching tasks in a far more effective manner. In the manual-intensive workflow, most processes relied on email-based resume processing, monitoring job boards, and manual matching, which turned out to be time-consuming, leading to slow placements, human errors, and lost opportunities. The organization developed an AI-enabled recruitment automation platform along with Bacancy in order to address the strict demands of compliance and real-time as well as accurate job matching. They have automated multi-format resume processing, job scraping, intelligent semantic candidate-to-job matching, and real-time reporting while cutting down human intervention, enhancing match accuracy, and hugely improving placement speed and efficiency in operations.

Technical Stack

  •  Python
  • FastAPI
  • Pandas
  • spaCy
  • PyMuPDF
  • PostgreSQL
  • Microsoft Graph API
  • Docker
  • AWS
  • Industry

    Healthcare

  • region
  • Region

    UK

  • project-size
  • Project Size

    Project Size Non-Disclosable

Highlights

100% automated resume processing and extraction

50+ daily job postings aggregated from multiple sources

AI-powered semantic matching improves accuracy by 85%

70% faster candidate-job placement with zero manual effort

Challenges & Solutions

Multi-Format Resume Processing and Text Extraction

  • Solution: It uses an AI-based multi-layered extraction system, where PyMuPDF is used along with python-docx and docx2txt, with OpenAI Vision API as the fallback OCR. This automatically detects poor extraction quality and switches to OCR for structured candidate data extraction from any resume format.

Job Scraping, Deduplication, and Rate Limiting

  • Solution: Bacancy implemented tailored AI automation expertise to build intelligent job scraping workflows that used unique job_id constraints, thread-safe scraping limits, and automated filters to capture only newly posted jobs. By incorporating connection pooling and locking mechanisms, the system eliminated duplicate records and ensured reliable performance during concurrent scraping operations.

Semantic Profession Matching with Varied Nomenclature

  • Solution: Developed an AI-driven profession normalization engine using GPT-4, which semantically mapped job titles like RN, Registered Nurse, and ICU Nurse to standardized professions. Applied intelligent caching, fallback strategies, and two-level matching for profession plus specialty and profession-only cases.

Real-Time Processing with Database Connection Reliability

  • Solution: Implemented a robust connection pooling and retry framework integrated with health monitoring, transaction management, and automated recovery to ensure uninterrupted email monitoring and candidate matching. The solution also included real-time logging and proactive alerting, enabling early issue detection, optimized performance, and 24/7 operational reliability.

Core Features

  • Continuous AI-based email monitoring and resume extraction
  • Automated multi-source job scraping and aggregation
  • GPT-4 and spaCy-based semantic candidate-job matching
  • Real-time FastAPI dashboard for job and candidate tracking
  • Automated Excel reports and email delivery
  • Intelligent caching, deduplication, and logging system
  • no.-of-resources
  • No. of Resources

    8

  • time-frame
  • Time Frame

    July 2024 to February 2025

Experience With Bacancy

2500+ Projects Experienced Innovation with Bacancy!

Get access to an experienced team of developers and engineers from Bacancy, handpicked to ace your goals. Kickstart within 48 hours, no-risk trial.

Book a 30 min call
14+

Years of Business
Experience

1458+

Happy
Customers

12+

Countries with
Happy Customers

1050+

Agile enabled
employees

How Can We Help?