Interfaces people enjoy. Systems they can trust.
I build across web, backend, AI and connected devices—with a focus on reliable workflows, clear user experiences and explainable decisions.
Projects · More repositories · About · Tech stack · Activity · Connect
Start here: my most substantial product, backend and AI work. Deployed projects include a live demo; local-only projects link to their setup instructions.
A complete booking experience, backed by an ownership-safe seat-allocation workflow.
- Searchable movie/live-event discovery, city selection, favourites and recommendations.
- Accessible seat maps, atomic holds, expiry, server-calculated totals and QR tickets.
- FIFO waitlists with single-use offers, cancellation and live inventory through SSE.
- Customer, organiser and administrator workflows, reports and audit history.
- Durable background jobs, retry scheduling and automated API/browser regression tests.
React Vite TypeScript Node.js Prisma PostgreSQL JWT Playwright Railway
▶ Live demo · Source code · Architecture & setup
Sample inventory and demo accounts—not real paid tickets. Email delivery remains pending HTTPS-provider integration; Railway Trial blocks SMTP.
Natural-language questions become bounded analytical plans, evidence and human-review workflows.
- Query-aware orchestration selects relevant tools and explains why others were skipped.
- Rules, statistics, Isolation Forest and graph analysis support explainable risk scoring.
- Governed datasets, model provenance, investigation traces and reviewer audit history.
- Human escalation remains explicit: investigative leads are not regulatory decisions.
Team project · My documented contribution: product experience, frontend architecture, UI/UX, visualization and integration.
React TypeScript FastAPI Python DuckDB scikit-learn NetworkX Groq
▶ Live demo · Source code · API docs
AWS deployment edition: Repository · Live AWS demo
Task planning, telemetry, safety alerts, incident handling and cited operator guidance in one prototype. Deterministic safety rules take priority over ML/LLM output; synthetic telemetry is labelled as demonstration data.
Python FastAPI PostgreSQL / SQLite Document RAG Docker
Source & local demo · Architecture
DepShield AI combines npm audit/OSV evidence, dependency topology, bounded static reachability, contextual risk, remediation simulation, policy gates and SBOM exports. Scores are project heuristics—not proof of exploitability or a security certification.
Next.js TypeScript SQLite OSV Gemini React Flow
Source & local demo · Architecture · Preserved project provenance
Live multi-user editing, formulas, presence, sharing, version history, in-document chat, charts and CSV export in a collaborative workspace.
Next.js TypeScript Firebase Firestore
| Project | What it explores | Links |
|---|---|---|
| Smart Helmet | ESP32/MPU6050 crash detection, Flutter monitoring, Firebase telemetry and GPS-tagged emergency alerts | Repository · Documentation |
| CampusMate | Flutter campus tours with Supabase auth, Google Maps, 360° panoramas, audio and 3D previews | Repository |
| Sentient Calendar | React/Next.js calendar, range selection, daily notes, sticky memos and local persistence | Live demo · Repository |
| Vivi Alarm | React productivity dashboard with clocks, weather, alarms and Pomodoro tools | Live demo · Repository |
| Online Voting System | Voter/candidate registration, voting confirmation and results with Node.js, Express and SQLite | Repository |
| Emergency Vehicle Detection | Python project exploring emergency-vehicle detection | Repository |
Browse all public repositories →
Featured work is curated rather than an exhaustive list of forks, experiments or private repositories. Local-only projects are labelled instead of given invented deployment links.
I work across frontend, backend and applied AI. I enjoy engineering problems where details matter: concurrent state changes, role-based access, database consistency, secure model integrations and recovery from partial failure.
- Currently building: full-stack products, explainable AI workflows and reliable backend systems.
- Interested in: API architecture, distributed state, system design, accessible interfaces and connected applications.
- Open to: Software Engineer Intern, Full-Stack, Backend and Applied AI/ML opportunities.
- Engineering mindset: Build → Test → Understand → Improve.
A polished interface matters. So do the behaviour under invalid input, unavailable services and simultaneous users—and the documentation that makes those tradeoffs clear.
Tools represented in my projects—not a list copied from a template.
Languages
Frontend & mobile
Backend & data
AI, testing & deployment
All cards request data for Mayank2142; no numbers are hard-coded. Third-party cards can be cached or unavailable and may not include all private activity. Language proportions describe repository code, not proficiency. The letter rank is the card provider's activity metric—not an official GitHub rating or a measure of engineering ability. See GitHub's native contribution calendar on my profile for the first-party activity view.
Interested in building a product, discussing systems or collaborating on applied AI? Message me on LinkedIn.
GitHub repositories · LeetCode · Codolio
Thanks for visiting — explore a demo, read the code, and let's build something useful.

