Projects

Built from scratch: a Meta ads automation system (Python + Google Apps Script), a verified MFD database (web scraping + AMFI ARN matching), an N8N workflow suite for marketing ops, and a 6-desk Claude AI editorial infrastructure. Total daily hours saved: 13–16. Annual tech cost saved: ₹5L+.

18-field CRM with a 4-stage pipeline, an AI lead scoring model with 6 weighted signals (Hot/Warm/Cold classification), automated follow-up reminders, and AI-generated personalised outreach drafts. Designed around how the sales team actually worked, not how HubSpot thought they should.

Full sentiment analysis and topic modelling across 1.3 lakh+ Google Play Store reviews from India's three biggest investment apps. VADER sentiment classification, LDA topic modelling, aspect-based sentiment scoring, and correspondence analysis. Translated findings into actionable marketing strategy.

Designed and developed the full ecommerce experience on WordPress for a new product category with near-zero market awareness. Product architecture built around use cases rather than specs. User journey designed to educate before selling, because the product required it.

Designed 30 structured digital training modules for UP Police personnel covering crowd management, first aid, fire safety, and emergency response. Built for comprehension under pressure, clear decision trees, short format, visual step-by-step. Cloud access system deployed for rapid field retrieval.

Built Power BI and Excel dashboards tracking application numbers, engagement rates, conversion trends, and CAC across 40,000+ lead database. Replaced manual weekly reporting with real-time visibility. Leadership could see what was working, and decisions got faster and better as a result.