About Me
Experienced software engineer with over 10+ years of expertise in building scalable systems, optimizing complex SQL/BigQuery queries, and leading high-performing teams. Passionate about AI integration, chatbot development, and query performance tuning. Adept at identifying technical bottlenecks and delivering innovative, cost-efficient solutions.
Technical Skills
Languages & Frameworks
JavaScript, Node.js, Express.js, Flask, Python, Core PHP, CodeIgniter, Laravel
Databases & Storage
MongoDB, PostgreSQL, BigQuery, ClickHouse, AWS DocumentDB, Materialized Views
Cloud & Infrastructure
Google Cloud Functions, Google Cloud Scheduler, AWS Lambda, RabbitMQ, Redis
Data & Analytics
Google Data Studio, Query Performance Tuning, Data Pipelines, Real-Time Data Processing, Data Engineering, Data Visualization, Business Intelligence
AI & Automation
OpenAI GPT-4o-mini, LLM Integration, Prompt Engineering, Customer Support AI Chatbot, Automation
Security & Identity
OpenID Connect, JWT, OAuth 2.0, PII Redaction, RBAC
Architecture & Patterns
Microservices, API Gateway, Event-Driven Architecture, Distributed Systems, Rate Limiting
DevOps & Monitoring
System Monitoring, Logrotate, Hatchet, MongoAnalyzer, Deadmansnitch, Cost Optimization, Anomaly Detection
Experience
Staff SDE at GoodGlammGroup
Apr, 2023 – Mar, 2025- Evaluated AWS DocumentDB vs MongoDB Atlas, benchmarking real-world workloads and uncovering high I/O costs under Standard storage—advised postponing migration until I/O-Optimized mode matured
- Designed advanced rate-limiting for critical APIs using Redis and context-aware identifiers (beyond IP) to prevent abuse without exposing internal logic
- Built courier performance reports in Data Studio using BigQuery to calculate delays across delivery stages—helping the supply chain team optimize courier selection
- Automated discount abuse detection via a cron-triggered Python script analyzing orders-per-minute; sent real-time alerts using Deadmansnitch to prevent losses
- Refactored reporting to an event-driven model, incrementally aggregating order metrics in real-time and reducing DB load; evaluated PostgreSQL materialized views as an alternative
- Improved MongoDB performance by analyzing slow query logs (Hatchet, MongoAnalyzer), refining indexes based on ESR rules and validating improvements via Atlas metrics
- Developed an automated data sync script to migrate records from prod to pre-prod during off-peak hours with PII redaction for compliance
- Led BQML ARIMA PLUS POC to forecast warehouse stock, analyzing seasonality and stockout risks; after cost and accuracy review with Google, advised against production use
- Built data ingestion pipelines using Cloud Functions + Scheduler, fetching and transforming third-party API data (courier, warehouse) into BigQuery
- Developed an LLM assistant that converted natural language to BigQuery SQL using prompt engineering and schema context, enabling near-instant insights for leadership
- Built a GPT-4o-mini chatbot mapped to 20+ support use cases, integrated with backend APIs for real-time user/order/refund data—automated ∼70% queries, improving CSAT
- Reduced OpenAI token usage by building a use-case classifier that narrowed context passed to the LLM, cutting input tokens by ∼10× and reducing API costs
- Resolved a critical disk alert loop by analyzing log patterns and implementing logrotate on the data engineering server—eliminating recurring 100% disk space usage issues
- Created materialized views over 2TB+ clickstream data in BigQuery using partitioning and clustering to accelerate queries for data science and analytics teams
- Optimized ClickHouse performance by introducing intermediate materialized tables and improving aggregation logic—cut compute costs and reduced redundant work
- Built a GCS → BigQuery ingestion pipeline, transforming vendor data from external tables and aggregating results using materialized views for downstream access
- Developed real-time anomaly alerts for order volume and regional behavior using BigQuery thresholds—flagged potential abuse and operational issues early
- Identified traffic flagged as attacks as user behaviour misuse, not a true threat; implemented rate-limiting to reduce abuse while preserving UX for valid users
- Automated vendor invoice parsing with a Python serverless app that extracted PDFs from emails and converted them into structured Sheets using Gemini APIs
Software Engineer at MyGlamm
Jun, 2017 – Mar, 2023- Built the first Payment Microservice alongside a legacy monolith to support full payment and refund flows with essential transaction logging
- Automated daily order and revenue reports for each franchise using Core PHP and Mailchimp, improving reporting accuracy and reducing manual effort
- Secured social login integration in an Express.js API Gateway by verifying Google/Facebook JWT tokens at the backend, ensuring safe user onboarding
- Replaced MongoDB reporting with asynchronous BigQuery jobs, reducing production DB load and accelerating large-scale report generation
- Cut BigQuery costs by 4× via partitioning, clustering, and authorized views; later switched to flat-rate pricing for predictable cost control
- Designed real-time config propagation using RabbitMQ fan-out and Redis to reflect config updates instantly across all services without redeployments
- Led a POC for OpenID Connect-based Auth microservice, evaluated impact on UX/conversion, and advised against deployment due to business concerns
- Built a Flask service to analyze order patterns and delivery outcomes, generating risk scores to improve fraud detection and operational decisions
- Co-designed a performant URL shortener, optimizing key length and evaluated Bloom filters as an addon for fast, collision-resistant URL generation
- Conducted a MongoDB sharding POC, exploring shard key strategies and re-sharding mechanisms to scale high-volume data workloads efficiently
Software Engineer at B2 Information Systems Pvt. Ltd.
Mar, 2013 – Jul, 2015- Developed core features for client-facing projects including Ottawa Dental Society, MProfit, and the BiteBank Admin Panel, contributing to robust and maintainable codebases
- Progressed from core developer to project lead, consistently delivering high-quality releases on schedule and mentoring junior team members
- Led the migration of legacy BiteBank systems to a modern web framework, streamlining workflows and improving system reliability and maintainability
Education
MMS Operations
2015 – 2017Welingkar Institute of Management Development and Research, Mumbai
MSc Information Technology
2010 – 2012University of Mumbai, Mumbai
BSc Computer Science
2007 – 2010University of Mumbai, Mumbai
Open Source Contributions
node-time-sleep
Python-like time.sleep(ms) for Node.js with non-blocking async/await support.
json-cache-key-generator
Deterministically generates a cache key from deeply nested JSON — useful for caching and memoization.
remove-invisible-characters
Removes ASCII extended characters that are invisible but can break Node.js code — ideal for input sanitization.
Side Projects
Global Mute for Google Meet
Chrome Extension to mute/unmute Google Meet globally using hotkeys — works even outside the browser.
Available on Chrome Web Store →Skip Silence (In Progress)
Chrome Extension that detects near-silent moments in YouTube videos and speeds up playback to 2× during silence. Returns to normal speed once speech resumes — ideal for long lectures. Supports real-time audio analysis and customizable fast-forward settings.
Language Learner (German)
Helps users learn German by showing translated sentences from English, Hindi, or Marathi and assessing pronunciation accuracy. Uses 4 transcription methods: SpeechRecognition interface of the Google Web Speech API, OpenAI Whisper (API & FOSS), Azure Speech-to-Text. Provides feedback on pronunciation with visual improvement graphs over multiple attempts.