open to internships · Mumbai, India
Krish Mistry

Krish Mistry.

I build things that actually ship — full-stack apps, AI-powered systems, and real-time platforms that solve real problems with clean, measurable code.

8 production apps
9.2 CGPA · SFIT
100+ participants led
2nd Colloquium '25
krish@portfolio ~ profile.json
KM
Krish H. Mistry
Full Stack · AI/ML · Mumbai
$cat profile.json
role: "Full Stack Developer"
institute: "SFIT, Mumbai"
cgpa: 9.2 (avg · BE IT)
batch: "2023–2027 · Final Year"
shipped: 8 production apps
led: 100+ event participants
hackathon: "2nd · Colloquium '25"
team: "Lead · Cipher Monks"
ideathon: "Organiser · Samadhan '26"
nss: "Head · SFIT"
chess: "national competitor"
$
primary stack
Next.js 15 React 18 TypeScript Python Django REST Node.js Flutter TensorFlow Gemini AI Supabase Firebase PostgreSQL WebRTC Socket.IO Docker Vercel
scroll

Built on curiosity,
shipped with intention.

I'm a final-year IT student at St. Francis Institute of Technology, Mumbai — but I think of myself as a product builder first, student second. Since my first "hello world" I've been obsessed with closing the gap between idea and working software.

Over the past two years I've shipped 8 production-grade apps — everything from a real-time geopolitical intelligence dashboard powered by Gemini AI, to a cross-platform community safety TensorFlow's MobileNetV3. I don't just build for portfolios; I build to solve real problems.

How I think when I'm not shipping.

Chess — National Competitor

The same long-horizon thinking that wins endgames shapes how I architect systems. Play a quick game below — you're White.

click a piece, then a square
Your move (White)
🧩
Rubik's Cube

Pattern recognition under time pressure — a habit that transfers directly to debugging. Drag to view, use the buttons to twist a layer.

drag to view

My Journey

Building software and building teams turned out to need the same skill: clear thinking under pressure.

Started Coding 2022
Self-taught · First lines of code
  • Wrote my first line of code and fell in love with building things from scratch — the moment that started everything.
B.E. Information Technology 2023 – 2027
St. Francis Institute of Technology Ongoing
  • First year dived deep into DSA, OS, and web fundamentals — the base everything else is built on.
First Full-Stack Projects 2024
EduCycle & HeritageGuide
  • Built real-world apps with full backends, interactive maps, and API integrations — working alongside other contributors for the first time.
  • What it taught me: managing Git workflows, code reviews, and feature splits across teammates with different strengths and schedules.
⚙ cross-functional teamwork
Head of NSS 2025
SFIT · National Service Scheme
  • Took on community outreach leadership — translating program goals into clear action plans for 50+ volunteers.
  • What it taught me: writing a clean commit message and briefing a volunteer team are the same skill — say exactly what needs to happen and why.
💬 clear communication
Hackathons, Ideathons & Ambassadorship 2025 – 2026
Colloquium '25 · Samadhan '26 · Google Current
  • Runner-up at Colloquium '25, SFIT's national-level hackathon — then organized Samadhan '26, a national ideathon with 100+ participants and NGO partnerships, from scratch: venue, timeline, sponsor outreach, and a day-of run with zero critical failures.
  • Selected as a Google Student Ambassador soon after — all while maintaining a 9.2 CGPA and shipping production apps in parallel. Not by working harder, but by knowing what to cut, defer, and finish first.
🎯 leadership under pressure ⏱ time management

Currently Building.

AI Web Development Intern

@ InAmigos Foundation

Developing responsive web applications, integrating AI functionalities, and optimizing user interfaces for performance.

Completed • June - July

Machine Learning Intern

@ Flyrank

Developing predictive models, analyzing large datasets, and deploying ML pipelines.

Active • Since July

Google Student Ambassador

@ Google

Advocating for Google technologies, organizing community events, and mentoring students.

Active • Since June

Where I've worked & competed.

Click a card to flip it and see the details.


Google Student Ambassador
Google Student Ambassador
@ Google · June 2026 – Present
tap to flip ↻
Google Student Ambassador
@ Google · June 2026 – Present
  • Representing Google on campus, acting as a liaison between the university and Google to foster a vibrant developer community.
  • Organizing workshops, tech talks, and hackathons on Google Cloud, Android, and AI/ML.
  • Collaborating with other student leaders to improve the technical ecosystem at the university.
Community BuildingLeadershipDev Relations
InAmigos Foundation
AI Web Development Intern
@ InAmigos Foundation
tap to flip ↻
AI Web Development Intern
@ InAmigos Foundation · June – July 2026
  • Building and maintaining AI-integrated web features for a social impact platform, focused on performance and accessibility.
  • Shipping production-ready components with the engineering team for underserved communities.
AI IntegrationWeb DevRemote
Flyrank
Machine Learning Intern
@ Flyrank
tap to flip ↻
Machine Learning Intern
@ Flyrank · July 2026 – Present
  • Designing and building scalable ML models for Flyrank's core product, focused on accuracy and performance.
  • Owning end-to-end ML feature delivery — from data processing to production deployment.
Machine LearningAI ModelsPython
Cipher Monks
Team Leader
@ Cipher Monks
tap to flip ↻
Team Leader
@ Cipher Monks · June 2026 – Present
  • Spearheading a dedicated engineering team across 2–3 major hackathons under tight deadlines.
  • Driving system architecture, delegating tasks, and ensuring end-to-end execution.
  • Mentoring teammates and facilitating effective communication.
LeadershipHackathonsSystem Design

The stack I reach for.

Grouped by domain — depth over breadth, always.

Languages
C C++ Java JavaScript TypeScript Python Dart
Frontend & Mobile
Next.js Flutter Angular Angular.js Three.js TailwindCSS HTML5
Cloud & DevOps
Firebase Vercel AWS Google Cloud Docker Git
AI & Data Science
TensorFlow Pandas NumPy Matplotlib
Backend
Node.js Django Express.js Flask
Databases
MySQL SQLite Supabase MongoDB Postgres

Technical Expertise

I design and build scalable, high-performance web applications from idea to production — with a strong focus on usability, performance, and clean architecture.

What I Do

Full-Stack Web Development
AI & Machine Learning
Mobile App Development
Backend Architecture & APIs
Cloud & DevOps
Interactive 3D & Data Viz

Things I've shipped.

Real projects — not tutorials. Every one of these is production-grade and deployed or deployable.

ImpactGlobe Dashboard

ImpactGlobe

Live Web ↗
For traders & analysts who need geopolitical context without digging through 20 news tabs — a live intelligence layer that maps world events directly to Forex impact.
  • 3D globe (Three.js + Leaflet) mapping live geopolitical events with severity tiers (Critical, High, Medium) and historical 24-hour replay.
  • Gemini 2.0 Flash auto-generates structured news intelligence with direct Forex impact analysis (e.g., EUR/USD, XAU/USD).
  • Real-time event tracking and live market panel powered by Next.js and Firebase — zero manual refresh needed.

⚙️ Hardest Technical Decision

The trickiest challenge was syncing live geopolitical event data with Three.js globe rendering without causing frame drops. Solved it by debouncing Firebase listeners and batching DOM updates outside the animation loop.

HemiSphere App UI

HemiSphere

Live Web ↗
Urban communities lack fast, low-cost tools to report and act on safety hazards. HemiSphere puts AI-powered incident classification and CO₂ tracking directly in residents' hands — no cloud dependency needed.
  • MobileNetV3 CNN (Transfer Learning) classifies accidents, construction hazards, and garbage from incident photos with on-device inference.
  • Heuristic regression for real-time CO₂ emissions tracking — gives communities actionable environmental data locally.
  • Flutter + Firebase Firestore backend with real-time sync across all community members.

⚙️ Hardest Technical Decision

The key decision was choosing on-device inference (MobileNetV3) over a cloud API — this eliminated latency, worked offline, and kept sensitive incident photos off external servers. The tradeoff was a larger app bundle, offset with lazy loading.

More shipped work.

Flip a card to see what's under the hood.

TransitOps
flip ↻TransitOps

Enterprise fleet & logistics command center.

TransitOps
  • Comprehensive logistics dashboard bridging hardware tracking, human resource management, and financial analytics.
  • Real-time trip lifecycle planning, fuel intelligence with anomaly detection, and driver operations tracking.
  • Role-Based Access Control with gated write operations, audit logging, and data export functionalities.

⚙️ Hardest Technical Decision

Building a fully synchronized, real-time interface across multiple modules (fleet, dispatch, fuel) required careful state management and real-time database architecture to prevent redundant fetches and ensure absolute data consistency.

React 19TypeScriptFirebaseTailwind
CrowdPulse
flip ↻CrowdPulse

Real-time Crowd Analysis Dashboard.

CrowdPulse
  • Engineered real-time object detection and density mapping pipelines using YOLOv8 to proactively identify crowd anomalies.
  • Developed a Flask web dashboard featuring passkey authentication and interactive, dynamic hotspot zones for targeted monitoring.
  • Architected and deployed the AI model on Hugging Face Spaces, utilizing PyTorch and CUDA for GPU-accelerated inference.

⚙️ Hardest Technical Decision

Balancing high-accuracy tracking with the low latency required for real-time alerts. I optimized the YOLOv8 tracking algorithms and moved all AI inference to an NVIDIA GPU using PyTorch (CUDA 12.1), successfully achieving real-time processing speeds without sacrificing precision.

PythonFlaskYOLOv8
AyurWell
flip ↻AyurWell

AI-generated diet plans & dosha assessment.

AyurWell
  • Formulated an intricate scoring algorithm mapping user survey inputs to precise Ayurvedic Dosha profiles.
  • Integrated LLMs to dynamically generate highly personalized, season-aware dietary plans.
  • Constructed a fluid dashboard using Recharts to visualize wellness progress over time.

⚙️ Hardest Technical Decision

Feeding complex health profiles into the LLM safely was critical. I implemented a strict prompt chaining pipeline with validation layers to ensure the AI generated safe advice.

Vite + ReactDjango RESTTailwind
ChessAnalyzer
flip ↻ChessAnalyzer

Browser-based Stockfish engine.

ChessAnalyzer
  • Play-vs-Stockfish with adjustable strength, multi-engine support with WASM threads gracefully degrading based on CPU.
  • Move classification, multi-PV engine lines, and game summaries highlighting inaccuracies, mistakes, and blunders.
  • 40+ piece themes, SAN move list, zero backend dependency for analysis.

⚙️ Hardest Technical Decision

The biggest challenge was running a heavy C++ chess engine in JS without freezing the UI. I compiled Stockfish to WebAssembly (WASM) and ran it inside Web Workers, communicating via message passing.

Next.js 15TypeScriptStockfish WASM
HeritageGuide
flip ↻HeritageGuide

Interactive map with routing.

HeritageGuide
  • Dynamic map integration using Leaflet and OpenStreetMap for accurate site discovery.
  • Built a full-stack Django architecture handling secure slot booking, authentication, and routing.
  • Firebase-backed features ensure real-time data flow for peak hour updates and reviews.

⚙️ Hardest Technical Decision

The routing engine needed to be fast and multimodal. Instead of building from scratch, I integrated with OpenRouteService APIs but cached frequent routes in Redis to minimize API costs and latency.

Django 5LeafletFirebase
EduCycle
flip ↻EduCycle

Verified peer-to-peer marketplace for students.

EduCycle
  • Integrated Stripe for secure, verified student-to-student transactions.
  • Developed a robust PWA architecture enabling students to browse and message offline on campus.
  • Utilized Redis caching to serve the marketplace feed in under 100ms.

⚙️ Hardest Technical Decision

Managing real-time P2P chat alongside transactional data required careful planning. I decoupled the chat microservice using WebSockets while Django handles heavy DB relations.

Django RESTReact.jsStripeRedis
WatchParty
flip ↻WatchParty

Real-time synchronized video streaming.

WatchParty
  • Engineered a custom WebRTC signaling server using Socket.IO to establish low-latency peer connections.
  • Implemented robust timestamp synchronization algorithms to ensure sub-second video playback alignment.
  • Built lightweight, zero-install architecture focusing on pure client-side P2P media flow.

⚙️ Hardest Technical Decision

Instead of relaying video through a central server (expensive and slow), I forced a pure P2P mesh network for smaller rooms. The signaling server purely orchestrates connections.

WebRTCSocket.IONode.js
Nexus Financial
flip ↻Nexus Financial

High-performance fintech dashboard.

Nexus Financial
  • Architected a double-entry ledger system in Java to guarantee atomic, ACID-compliant transactions.
  • Enforced strict security measures using JDBC prepared statements to prevent any SQL injection vulnerabilities.
  • Developed a comprehensive dual-table audit trail to track every single balance change and user action.

⚙️ Hardest Technical Decision

Ensuring transaction atomicity during concurrent ATM operations was paramount. I heavily utilized MySQL row-level locking (SELECT ... FOR UPDATE) to prevent race conditions.

Java 22MySQL 8JDBC

GitHub & LeetCode Activity.

GitHub Streak
KrishMistry18's LeetCode Stats

GitHub Heatmap

KrishMistry18's Github heatmap

LeetCode Heatmap

LeetCode Heatmap
Activity Graph

TROPHY ROOM

Achievements and certifications unlocked


Let's build
something great.

I'm actively looking for internship opportunities in full-stack, AI/ML, or product engineering. If you're working on something interesting — or just want to talk tech — reach out.

say hello ↗