Sparsh Verma
Data Engineer & Cloud Automation Specialist
I design and deliver resilient data pipelines and cloud-native automation on GCP — helping teams replace slow, manual reporting with systems that run themselves. 3 years of production delivery across enterprise clients and personal products.
Background
I'm a data engineer currently at HCLTech, where I lead the design and delivery of cloud-native data pipelines and automation tooling for enterprise clients. My work spans GCP infrastructure (BigQuery, Cloud Run, Terraform), Python ETL, applied LLM pipelines, and end-to-end BI delivery.
Before HCLTech, I built BI dashboards and data pipelines at Blackcoffer Consulting, where I worked directly with clients across sectors to shorten the path from raw data to decision.
Alongside enterprise work, I independently built and launched SSBForge, a live web platform serving India's defence aspirants — giving me full ownership from backend architecture (Supabase, GitHub Actions CI/CD) to product design and growth.
- Location
- Delhi NCR, India
- Education
- B.Tech Computer Science
Galgotias College of Engineering, 2019–2023 - Focus areas
- Cloud pipelines · BI platforms · Applied GenAI
- Availability
- Open to new engagements
Technical Expertise
Certified across Google Cloud and AWS, with hands-on production delivery across the full data stack.
Professional Experience
Software Engineer
HCLTech · Noida, India
- → Led a 7-person team building a deterministic AI/LLM pipeline that replaced a fully manual content-review step, achieving 99% PII redaction accuracy.
- → Automated Python → BigQuery → Power BI reporting pipelines for enterprise clients, reducing manual reporting effort by 35%.
- → Built and shipped a PowerApps / Power Automate / FastAPI survey tool to 10,000+ employees via SMTP automation, cutting manual effort by 60%.
- → Engineered CI/CD workflows for GCP Cloud Run deployments, reducing deployment time by 60%.
- → Implemented Terraform-based GCP disaster recovery, enabling one-click infrastructure rebuild and 100% business continuity.
Data Science Intern
Blackcoffer Consulting · Delhi, India
- → Built FastAPI pipelines connecting Looker Studio to BigQuery, improving client reporting speed by 30%.
- → Designed and published interactive Power BI / DAX dashboards for client stakeholders across sectors.
- → Automated Python ETL scripts for data ingestion from external APIs (RapidAPI, Looker API).
- → Built web-scraping workflows (BeautifulSoup) feeding BigQuery and the ELK Stack for ongoing analysis.
Selected Work
Six production builds — one shipped solo end-to-end, the rest delivering measurable outcomes for enterprise clients.
SSBForge — Defence Exam Prep Platform
Live · Personal ProjectA full-stack web platform providing free timed practice for India's SSB defence interview — six modules (TAT, WAT, PPDT, SRT, Lecturette, GPE) with Supabase auth, Razorpay premium tier, and GitHub Actions CI/CD. Designed, built, and launched independently.
Forge Your Officer's Mindset
Stack
Highlights
- · Supabase Auth with session caching, token refresh, and RLS-controlled signed URLs for premium content gating.
- · GitHub Actions CI/CD injects build-time secrets, keeping the repo public without exposing premium assets.
- · Fully self-directed — sole engineer, designer, and product owner from zero to live.
AI / LLM PII Redaction Pipeline
HCLTech — ConfidentialA deterministic hybrid pipeline combining regex rule-matching with LLM verification (LangChain + OpenAI / Ollama) to scrub PII, profanity, and sensitive content from text at scale — replacing a fully manual review step for a 7-person team. Identical input always yields identical redactions, which is critical for compliance.
Stack
Highlights
- · 99% redaction accuracy across PII categories.
- · Deterministic by design — same input always yields same redaction output (compliance-safe).
- · Freed a 7-person team from manual review; shipped as a FastAPI service consumed by downstream workflows.
← swipe to see the full pipeline →
Client Data Automation — GCP & Power BI
HCLTech — ConfidentialEnd-to-end automated reporting pipelines for enterprise clients: SharePoint API → Cloud Run ETL → BigQuery → Power BI, with Terraform-managed infrastructure and GCP Cloud Build CI/CD.
Stack
Highlights
- · Reduced manual reporting effort by 35% and data-freeze cycle from 4 hours to 30 minutes.
- · Deployment time cut 60% via Cloud Build CI/CD; Terraform enables one-click infrastructure rebuild.
← swipe to see the full pipeline →
Interactive BI Dashboard — Crimes in India
Live DemoA Looker Studio report fed by an automated GCP Python pipeline, surfacing NCRB crime data (2020–2022) across Indian states, age groups, and crime categories. No manual refresh — updated automatically.
UK Government Data ETL
BlackcofferPython ETL pipeline ingesting and refining large public UK government datasets for a client-facing web application backed by PostgreSQL.
← swipe to see the full pipeline →
Content Sentiment Pipeline
BlackcofferAutomated scraping and NLP sentiment scoring across 200+ websites on a scheduled basis, feeding a continuous analysis pipeline.
← swipe to see the full pipeline →
Client Web Development — CFG & Sura Bakehouse
Live · Client ProjectsTwo production websites built end-to-end for real clients in Lucknow — a combat sports gym and a cloud kitchen bakery. Both include custom booking / order flows wired to Google Apps Script backends, the same stack used here on this portfolio's contact form.
Combat Fitness Garage
- · Full single-page site: fight-night aesthetic, bold typography, animated ticker, stats bar (380 fighters, 47 medals, 12 state champions).
- · Booking form with three modes — Free Demo, Buy a Package, Pay Fees — wired to Google Apps Script + Razorpay payment hooks.
- · Five thematic sections with scroll-reveal animations; fully mobile-responsive.
Sura Bakehouse
- · Multi-page site (main + gallery); maroon & cream palette drawn from the owner's brand logo.
- · Order form with Google Apps Script backend — emails the chef and logs every order to a Google Sheet with a blocked-dates calendar to manage availability.
- · Customer review submission with moderation, custom-request form, and owner-editable menu toggle — designed for a non-technical owner.
Get in Touch
Whether it's a data engineering project, a consultation, or exploring what's possible — send a message and I'll respond within 24 hours.
Response time
Open to
- · Data Engineering engagements
- · Cloud Architecture consulting
- · GenAI / Automation projects
- · Full-time roles