Case study · 2024 — Present

Datasprings — Data Automation & API Platform

Django + Neo4j employment platform with scraping pipelines, RAG-based search, and cloud-native deployment.

Role: Software Engineer — Backend / Product Owner
Company: NEWPRODATA
Year: 2024 — Present
Status: Shipped

01 — Problem

The employment data space is fragmented — dozens of sources, inconsistent formats, stale records, and no trustworthy verification layer. Recruiters and platform users were blocked on slow, manual reconciliation.

02 — Approach

Designed a Django monolith that orchestrates scraping, normalization, verification, and graph-backed search. Built Selenium + BeautifulSoup pipelines feeding a cleaning/validation layer before writing to PostgreSQL for relational data and Neo4j / AWS Neptune for relationship queries. Wrapped everything in a REST API with JWT + SSO. Added a Retrieval-Augmented Generation layer over Bedrock so users get semantic answers instead of keyword hits. Deployed on EC2 behind Nginx + uWSGI with S3 for blobs, Lambda for async work, CloudWatch for observability.

03 — Outcome

Shipped a platform that replaced manual workflows and is now the team's source of truth. Owned every on-call incident solo — kept uptime stable while shipping new features in parallel.

◆
Zero-downtime deploys across 12+ production releases
◆
Replaced ~70% of manual verification work with automation
◆
Sub-second semantic search over the graph
◆
End-to-end ownership of 3+ parallel systems

04 — Tech stack

Django Python PostgreSQL Neo4j AWS Neptune AWS EC2 AWS S3 AWS Lambda Nginx uWSGI Selenium BeautifulSoup RAG

More work

2026 — Present

BleBluBla — An Open People Graph

Social-professional platform where candidates, businesses, and the terms connecting them form one living graph.

Read the case

2025 — Present

PLOS — Personal Life OS

A single-operator Django + Flutter platform that treats ten dimensions of a human life as one coherent system.

Read the case