Selected work

Projects that shipped.

Websites, mobile apps, AI, and data — every project here was built for real users or real analysis, not for a tutorial.

Kwigo.co — Travel Platform Screenshot
Client · Travel

Kwigo — Travel Platform Redesign

Increased WhatsApp Enquiries Higher booking conversion achieved within the first week of launch
Problem

Dated booking experience, weak brand positioning, and low enquiry conversion for Pune-based group trips.

Approach

Full redesign, stronger brand identity, and a streamlined booking flow built specifically for Indian travellers.

Result

Substantial boost in direct WhatsApp enquiries within week 1 of going live.

HTML/CSS/JSUI RedesignResponsive
AI · Mobile App

Sakhi — AI Companion App

100% Offline-Capable Privacy-first wellbeing app running local LLMs on-device with zero cloud dependency
Problem

Mental-wellness support that works offline, privately, and in Hinglish without transmitting data to cloud servers.

Approach

React Native app running a local LLM on-device via Ollama; mood tracking, journaling, breathing exercises, and verified India crisis helplines.

Result

Fully offline-capable, privacy-first companion app with zero latency and zero data exposure.

React NativeExpoOllamaLocal LLM
Trek Captain — Group Trek Management Dashboard
Product · Web App

Trek Captain — Trek Management App

70% Faster Admin Automated participant rosters & WhatsApp group itinerary generation
Problem

Manual participant tracking, chaotic payment collection (UPI/bank/cash), and tedious daily itinerary formatting for trek leaders.

Approach

Built a responsive Next.js SaaS dashboard with automated roster tracking, payment verification, and 1-click WhatsApp itinerary exports.

Result

Reduced trip administration time by 70% with instant automated roster exports.

Next.jsReactTailwind CSSVercel
Data · Uber x upGrad

Traffic Volume Forecasting

Dynamic Surge Insights Time-series forecasting model identifying peak urban congestion windows
Problem

Predicting peak hourly congestion patterns from raw historical traffic sensor data to optimize fleet dispatch and surge pricing.

Approach

Applied Pandas data cleaning, time-based feature engineering, and predictive modeling during the Uber x upGrad mentorship program.

Result

Surfaced actionable peak traffic patterns supporting dynamic driver allocation strategy.

PythonPandasTime SeriesMatplotlib
Data · SQL

IMDB Movie Analysis & Strategy

Studio ROI Framework Complex SQL relational analysis converting raw movie metadata into release strategies
Problem

Determining profitable release windows, director pairings, and genre investments for a studio planning global releases.

Approach

Designed a 6-table relational schema from raw IMDB data and executed complex SQL queries using CTEs, window functions (RANK, LEAD), and multi-joins.

Result

Delivered concrete, data-backed strategic recommendations for film greenlighting and marketing windows.

SQLMySQLCTEs & Window FnsExcel
Internship · Analytics

Telecom Churn Analysis

High-Risk Customer Identification Exploratory data analysis isolating primary cancellation drivers and tenure drop-offs
Problem

High customer churn rates eroding recurring subscription revenue for a major telecom provider.

Approach

Performed comprehensive EDA on customer datasets to analyze contract lengths, payment methods, and demographic cohorts.

Result

Segmented high-risk customer groups and flagged key pricing friction points to enable targeted retention campaigns.

PythonPandasEDAMatplotlib
Want something like this built? Explore services