AlokKulkarni

Software engineer. I build secure, real-time systems: networks, IoT, backends and apps. Final year, Computer Engineering, MIT-WPU Pune.

Projects

Atlas IoT, factory monitoring platform

Live readings from edge gateways become alarms, work orders, lockouts and OEE and MTTR reports for maintenance teams. Web, Electron desktop and Flutter apps run on one PostgreSQL backend, with tiered roles, an audit log and scoped API keys.

  • Next.js
  • TypeScript
  • PostgreSQL
  • Electron
  • Flutter
  1. Triage Agent on red: its evaluation output in a terminal

    Security Alert Triage Agent

    Investigates SIEM alerts on a free local model, cites evidence for every claim and waits for a human to approve any action.

    • Python
    • Ollama
    • MITRE ATT&CK
  2. Stockroom on navy: the dashboard and a product page

    Stockroom

    Inventory across warehouses and stores, with every movement in an append-only ledger. Stock can never go negative, and two people can't sell the last unit twice.

    • C#
    • ASP.NET Core
    • SQL Server
    • 29 xUnit tests
  3. Kharcha on blue: three app screens

    Kharcha

    Offline-first Android finance app with on-device AI: reads bank SMS, scans receipts, answers spending questions.

    • Flutter
    • On-device Gemma
  4. Torvik on black: the gear site and the live gear configurator

    Torvik Engineering

    Concept site for a gear maker with procedural three.js gears and scroll-driven motion, plus a live configurator that shapes gears in real time and exports STL, OBJ and GLB.

    • Astro
    • three.js
    • GSAP
  5. How late is the T? on green: the live dashboard, 64.6 percent on time

    How late is the T?

    Checks every Boston bus and train against the timetable every 30 minutes, with three layers of data-quality checks and a live dashboard.

    • Dagster
    • dbt
    • DuckDB
  6. Defect Inspection on yellow: the 15 part types it checks

    Visual Defect Inspection

    Finds defects on factory parts and shows where, with heatmaps, served by a FastAPI web service.

    • PyTorch
    • ONNX
    • FastAPI
  7. Smart Home Sensors on slate: the live sensor dashboard

    Smart Home Sensors

    Sensors over MQTT to a FastAPI backend with threshold alerts and a live WebSocket dashboard. One docker compose up.

    • MQTT
    • FastAPI
    • Docker
  8. Custom VPN on teal: cipher speeds on a Raspberry Pi

    Custom VPN with ML anomaly detection

    Capstone: a Layer-3 VPN on X25519 and ChaCha20 with no WireGuard underneath. Keys rotate every 2 minutes; I made throughput 5.5× faster by moving ML stats into their own process.

    • Python
    • Cryptography
    • Raspberry Pi
    • Android
  9. AR Anatomy Learning

    Team project: an AR app that puts 3D organs on a textbook page. I presented it and cut the project film.

    • Augmented reality
    • Team project

All repositories on GitHub

Experience

2024 – now

Freelance software developer

Self-employed, mobile and web · Pune

  • Delivered 3 Flutter apps for small businesses from the first client call to launch, owning the backend, database and hosting on most of them.
  • Built business websites in Next.js and Tailwind, including one for a manufacturing company, and wrote the copy where clients had none.
  • Built a company expense tracker in Flutter and Firebase where Gemini reads bills, GST lines included, with Firestore rules keeping each employee's data private.

2023 – 2027

B.Tech Computer Engineering

Dr. Vishwanath Karad MIT World Peace University · Pune

Coursework: computer networks, operating systems, DBMS, information and cyber security, cloud computing, big data, machine learning, compilers.

Skills

Languages
Python, C++, C, C#, TypeScript, JavaScript, Dart, Java, SQL, Bash
Cybersecurity
Firewalls (nftables, iptables), applied cryptography (X25519, ChaCha20, AES, TLS), key exchange and rotation, replay protection, ML anomaly detection on traffic, SIEM alert triage, threat intel lookups, MITRE ATT&CK, OWASP Top 10
Networking
TCP/IP, UDP, DNS, DHCP, routing and subnetting, packet capture and analysis (Wireshark, tcpdump), network traffic analysis, VPN tunnelling
Systems
Linux, shell scripting, systemd services, Raspberry Pi deployment
IoT and hardware
Raspberry Pi, ESP32, Arduino, GPIO, I2C, SPI, UART, sensors and actuators, MQTT (Mosquitto, QoS, topic design), edge gateways, real-time telemetry over WebSockets, time-series storage, threshold and offline alerting, industrial machine monitoring (OEE, MTTR)
Frontend and mobile
HTML, CSS, JavaScript (DOM, Fetch), React, Next.js, Tailwind CSS, Astro, three.js, GSAP, Flutter, Electron
Backend and APIs
.NET (ASP.NET Core MVC, Entity Framework Core, Razor), Node.js, Express, FastAPI, REST APIs, WebSockets
Databases and data
PostgreSQL, SQL Server, MySQL, MongoDB, Firebase, SQLite, DuckDB, Kafka, Apache Spark, Dagster, dbt
AI and ML
AI-assisted development (Claude Code), Gemini API, OpenAI API, Ollama, PyTorch, scikit-learn, pandas, MLflow, ONNX
DevOps and tools
CI/CD with GitHub Actions, unit and integration testing (xUnit, Vitest), Docker, Git, Android Studio
Film and creative
Cinematography on Sony ZV-E1, DaVinci Resolve editing and colour grading, motion design, DJing, piano

Also behind the camera

I shoot on a Sony ZV-E1 and edit and grade in DaVinci Resolve. Cinematography, colour grading, editing and motion design. I also DJ, play piano and train most days.

Marine Drive, Mumbai
Stills, at my desk
Training
Asakusa, Tokyo
Me facing the sea at sunset in Goa, straight off the camera in flat S-Log3 The same frame after my colour grade
Drag: raw vs graded, Goa
Tokyo at night
When was the last time you paused?
If you have no one to film, film yourself
My keyboard, from the film Pause
My piano

Let's build something.

alokkulkarni62@gmail.com

Pune, India