All case studies
AI-Powered Apps & Automation · Job-matching chatbot platform
Machine learning job recommendations
Replaced SQL-query job matching with a trained neural network, and cut database cost by moving older data to cloud storage.
- Industry
- Recruitment
- Service area
- AI-Powered Apps & Automation
The challenge
The platform connects job seekers and employers through a Facebook Messenger chatbot. Matching relied on traditional SQL queries, and database cost was rising every month.
What we built
- A deep neural network trained on the platform's historical job application data, recommending jobs to the seekers most likely to succeed.
- A data-reduction plan that shrank what the database stores, with older data moved to AWS S3 by Lambda functions.
The result
Job recommendations come from a learned model instead of fixed queries, and the data migration completed with no data loss.
Stack
- Python
- NumPy
- Pandas
- TensorFlow-Keras
- HDFS
- Node.js
- MongoDB
- AWS Lambda
- AWS S3
