Results
What I build
Scrapers and data pipelines
Clean, structured data from messy web sources. Scrapers that hold up on JavaScript-heavy sites, validation on every row, deduplication, and outputs ready to audit. Contact lookup APIs on top when you need them.
Browser automation and Chrome extensions
Work your team does by hand that a script could own. Browser bots, cloud-scheduled workflows and Chrome extensions, including rebuilding an existing extension to fit your process.
LLM integration
Models wired into a live product with structured outputs, cost controls and fallbacks. OpenAI and Gemini in a paying SaaS, and Qwen checking record quality across a 30M-record database.
Stack
Scraping and automation
Data and ETL
Cloud and DevOps
AI and LLMs
Backend and APIs
Frontend and storage
Projects in production
RadixSol — Healthcare data and vendor management
RadixSol · US remote · Jun 2026 – present
Problem Recruiters needed a clean, searchable base of healthcare professionals across US states, contact details for each person, and a way for staffing vendors to submit candidates against open jobs.
Approach Scrapers collect professional details. Python and PostgreSQL pipelines clean, deduplicate and validate 30M+ records, and a Qwen model flags inconsistent or suspicious ones. A lookup API returns contact details from a first name, last name and address. The vendor system at radixsolvms.com takes submissions against job postings synced through APIs.
BotRecruit — Healthcare shift automation
BotRecruit · UK remote · Jan 2025 – Jan 2026
Problem A UK healthcare platform booked nursing shifts by hand on JavaScript-rendered scheduling systems. It was error-prone, sensitive to time zones and impossible to scale.
Approach GCP-hosted Selenium and Playwright bots with session persistence and adaptive waits, containerized on GKE with cron-scheduled windows, automated validation and alerting.
VocalLocal — AI voice SaaS
CRM Group Consulting · US remote · Sep 2024 – Feb 2025
Problem The client needed multi-language audio transcription and translation, earning revenue from launch rather than shipping as a demo.
Approach Django and FastAPI backend on Firebase Firestore, GPT and Gemini for audio processing, JWT auth and Stripe subscription billing. Shipped to paying users.
India school intelligence dataset
Freelance · EdTech client
Problem An EdTech client needed normalized school data for all of India. Public sources were fragmented and needed automation at scale.
Approach A multithreaded scraping system with concurrent workers, retries with exponential backoff, schema-validated Pandas output and structured logging for recovery from failures.
400,000+ school records extracted, normalized and delivered
LLM PDF data extraction
Freelance · Self-hosted inference
Problem Standard PDF parsers failed on irregular layouts, and manual extraction from complex documents was an analytics bottleneck.
Approach Qwen-32B (Hugging Face) served on RunPod with OCR preprocessing, context-aware prompting for structured JSON and automated batch output.
Structured JSON from unstructured PDFs at batch scale
Google Maps business data scraper
Freelance · Multi-client deployment
Problem Clients needed enriched local business data for outreach campaigns, at scale, from a platform that blocks naive scrapers.
Approach A two-stage Playwright pipeline. Stage one discovers listings, stage two enriches each with contacts, ratings and reviews. Session management for anti-detection, schema-validated CSV output, reusable across campaigns.
80% outreach effectiveness increase through data quality
Work history
RadixSol — AI Engineer
US remoteOwns the data side of a healthcare staffing platform: scraping, cleaning, quality checks, lookup APIs and the vendor system recruiters work in.
BotRecruit — Software Engineer
UK remoteSole automation engineer for a UK healthcare staffing startup. Owned the full lifecycle from bot design to GCP deployment to production monitoring.
VocalLocal — Python Engineer
US remoteCo-engineered VocalLocal from zero to a revenue-generating AI SaaS with paying users. Built the backend, API layer, LLM integrations and billing.
Freelance Data Engineer
RemoteIndependent practice delivering automation and data pipelines for startups and agencies across India, the US and the UK.
How I work
Automation that runs without you
Every system I ship has cloud scheduling, retry logic, error alerting and validation built in, so nobody babysits a cron job at 2am.
Data you can decide with
Not just a CSV. Validated, schema-normalized data with audit logs, deduplication reports and freshness signals.
Real products, not side projects
VocalLocal has paying users. BotRecruit's scheduler runs production shifts. These systems carry real stakes.
Founding engineer mindset
I push back on scope, ask why before how, and care whether the system solves the problem, not only whether the ticket is closed.
Credentials
- B.Tech CSE, Parul University, Gujarat (2021–2025), CGPA 8.17
- SQL Bootcamp, Udemy · Certificate ↗
- SQL Intermediate, HackerRank · Certificate ↗
- GitHub Actions for CI/CD, LinkedIn Learning · Certificate ↗
- Introduction to Agent Skills, Anthropic · Certificate ↗
- Web Scraping Intern, JustGetLeads (Oct–Dec 2024)
- Data Engineer Intern, 1.5 Degree (Feb–Apr 2025)
Start a project
Tell me the site you need scraped, the workflow to automate or the LLM integration to ship. I'll scope it within 24 hours and deliver a working proof of concept within a week.
Freelance contracts · remote roles · startup engagements