Curiosity may start the project, but consistency finishes it. This is the daily rhythm behind the work, then the longer trail of streaks, experiments, and code.
Contribution calendar from 2025-10-19 to 2026-10-23: 283 of 370 days recorded activity, peaking at 55 contributions on 2026-07-15.
I learn by asking how things work. At Flo Mattress I used that habit to build the internal operating system the company runs on. I replaced 8+ outsourced tools with it at one-eighth of the cost, and teams across the company work inside it.
Each time I use software, I want to understand the decisions behind it and find the part I wish it did better. I look at a company's processes the same way: I put dashboards and logs on them, watch where work slows down or breaks, then rebuild that step and measure the result.
If other people face the same friction, I build the smallest useful version, use it myself and rebuild what is missing, in reusable parts, so a new integration takes one adapter. I pick up in days what takes months to build, and I try new AI models and tools as they appear.
Today I lead engineering at Flo Mattress as a founding member of the team. I gather requirements from the teams that use the system, size the load and the running cost before I design anything, and plan for what will fail. I set the coding standards, including the AI-assisted workflow, mentor the engineers and interns who join, and can lead several teams. The scale is larger and the habit is the same: question the familiar, connect the pieces and build the version I believe should exist.
Full-stack products with modern frameworks and cloud infrastructure.
Machine learning, deep learning, and analytics systems in production.
Scalable data infrastructure, pipelines, and storage.
Dashboards and decisions from raw data.
Education grounded my curiosity in information technology, computer science, and data—and taught me to keep learning beyond the syllabus.
Mumbai University
Data science, machine learning, deep learning, AI agents, business intelligence, social analytics, and time-series analysis.
Rizvi College of Arts, Science and Commerce
Focused on web services and business-intelligence projects.
Most of these began with one of two questions: how does this work, or what is it missing? These are the answers I turned into products.
Platform · Architecture
A multi-service platform with a Go auth gateway (rotating refresh tokens, HMAC-signed internal identity), FastAPI AI services and a React 19 dashboard, built with per-service Docker images.
Stack
Applied AI · Agents
A LangGraph and FAISS pipeline that routes content by persona, drafts with search-augmented generation and provider fallbacks, and defends conversations against prompt injection.
Stack
Applied AI · Careers
An AI-assisted resume reviewer with tailored evaluation, keyword analysis, ATS feedback, and cover-letter generation.
Stack
The finished projects are only part of the story. These pinned repositories keep the experiments, self-built versions, utilities, and questions that led to them.
Minimal, self-hosted URL shortener built with Go, React, and PostgreSQL
RAG-based cognitive bot loop: route, generate, and defend with persona-based arguments
Secure disposable/temporary email service with real-time inbox and dark mode
Disappear on camera using hand gestures — no green screen needed
Agricultural intelligence platform for Indian farmers with AI chat, crop management, and market price analysis
Portfolio of data science projects: ML, NLP, analytics, healthcare, and market research
Curiosity also sends me into other people's code. These are the fixes and improvements that earned their place in the systems where they began.
Building answers the first question; writing forces me to understand the answer. These notes preserve the decisions, failures, and patterns worth carrying forward.
A practical introduction to descriptive statistics, probability, inference, regression, and Bayesian reasoning with Python examples.
Read on site
Why deploying and maintaining machine-learning models requires the shared practices of data science and DevOps.
A practical tour of sales, customers, inventory, marketing, recommendations, fraud, and supply-chain analytics.
The technical work, collaboration, reporting, and communication that shape a data scientist’s working day.
A guide to statistical, predictive, forward-fill, and backward-fill approaches for handling missing observations.
If you are solving a problem, exploring an idea, or rebuilding something that could work better, I would like to hear about it.
// Timezone IST (UTC+5:30). Cold outreach is welcome if it is specific.
exit 0 — that is the whole page. Thanks for reading to the bottom.
EOF
