I am

MANJIT KUMAR DAS

B.Tech (CSE)

Manjit

Hi, I'm Manjit Kumar Das

I'm a B.Tech student at Dhemaji Engineering College with a deep passion for computer science and software development. I thrive on tackling complex problems and turning them into clean, efficient code.

B.Tech Dhemaji Engineering College 2028
H.S Sigma Point Integrated Academy 2024
H.S.L.C St. Xavier's High School 2022

Skills

  • Python
  • C
  • C++
  • Java
  • HTML
  • CSS
  • Javscript
  • React
  • Node Js
  • npm
  • MySQL
  • MongoDB
  • Github
  • Data Structures

Projects

Some are live, some are still on my machine.

Marg project screenshot

Marg (The Path)

Live

A career guidance platform that helps students explore academic streams, discover 25 detailed career profiles (with key skills and outlook), and browse internships — all with instant search.

  • React
  • Node.js
  • Firebase

Marg (The Path)

Overview

Marg (The Path) is a career guidance web platform for students. Most job portals list vacancies but never explain how to get there, so Marg helps 10th-pass, 12th-pass and college students discover career options, understand the skills each one needs, and browse internships and jobs. Students explore Science, Arts and Commerce streams, read 25 career profiles, and use instant search to find what they need. It was built by a team of five as our 4th-semester B.Tech CSE micro project at Dhemaji Engineering College. It is a React single-page application with Firebase authentication, deployed on Vercel.

Key features

  • Stream-wise exploration: browse Science, Arts and Commerce, each with its own set of career paths.
  • 25 career profiles: each has an overview, the key skills required and the industry outlook (will add more)
  • Instant search across careers and internships
  • Jobs and internships board: search plus Internship, Full-Time and Remote filters.
  • Secure sign-in: email and password or one-click Google sign-in through Firebase Authentication, plus password reset by email.
  • Responsive design: works on desktop and mobile.
  • Single-page experience: pages change without a full reload.

Tech stack

  • React
  • Node.js
  • Firebase
Live demo Code
MAYA project screenshot

MAYA (Intelligence Core)

Not deployed

A voice-controlled desktop AI assistant with a futuristic Arc Reactor GUI. Built with CustomTkinter, featuring speech recognition, text-to-speech, fuzzy commands, system monitoring, and local app control.

  • Python

MAYA (Intelligence Core)

Overview

M.A.Y.A. (Intelligence Core) is a custom-built, voice-controlled desktop AI assistant designed with a futuristic aesthetic inspired by Jarvis. Developed as a personal project, it combines a sleek dark-themed GUI with powerful local system control and natural language interaction.

Key features

  • Futuristic Arc Reactor GUI:

    A custom-designed interface featuring a dynamic, animated Arc Reactor core at the center. The reactor changes color and rotation speed based on the assistant's state — idle (blue), listening (red), or speaking (bright cyan) — creating an immersive visual experience.

  • Voice Interaction:

    Full speech recognition and text-to-speech capabilities. Maya listens for voice commands through the microphone and responds using a natural-sounding female voice. Ambient noise adjustment ensures better accuracy in real-world environments.

  • Fuzzy Command Matching:

    Uses intelligent fuzzy string matching so commands don't need to be spoken perfectly. Phrases like “hey Maya”, “what's the time”, or “open visual studio” are recognized even with slight variations.

  • System Monitoring:

    Instantly reports real-time system status including CPU usage and battery percentage when asked.

  • Application & File Control:

    Can open frequently used applications such as:

    - Visual Studio Code

    - Dec-C++

    - Notepad

    - And more (that needs to be added)

    Also supports force-closing applications by name

  • Web Integration:

    Opens websites and perform searches on command:

    - Google

    - Youtube

    - Youtube Music

    - Custom Google searches

  • Personalized Responses:

    Recognizes its creator and provides contextual replies. Includes special responses for identity questions and can open the developer's Instagram profile on request.

  • Live Terminal Log:

    A glassmorphism-style output console displays every interaction in real time — user commands, system status, and Maya's responses — for transparency and debugging.

  • Multithreaded Architecture:

    Voice processing runs in a separate thread so the GUI remains smooth and responsive during listening and speaking operations.

    • M.A.Y.A. is a fully offline-capable personal assistant that blends visual appeal with practical desktop automation.

Tech stack

  • Python
  • CustomTkinter
  • SpeechRecognition
  • pyttsx3
  • psutil
  • thefuzz
Code
Disha project screenshot

Disha (Direction)

Live

A multilingual, voice-enabled web platform that helps parents and learners explore skill-based careers together, with outcome data, a cost-and-earnings explainer and human counsellor support.

  • React
  • Node.js
  • Firebase

Disha (Direction)

Overview

Disha (Direction) is a full-stack-style web platform built for Smart India Hackathon 2026. It tackles the parental perception gap in vocational training. Many learners would do well in skilled trades, but parents often see these careers as low status, poorly paid or unsafe, and most career tools only talk to the student.
Disha brings parents and learners into one shared counselling experience. A joint AI chat lets both sides ask questions in their own language, by typing or by voice. Every concern is answered with evidence, such as salary, placement and NSQF progression paths, instead of generic advice. A cost-and-earnings explainer compares vocational and degree routes for the family's own context. If the AI isn't enough, the family can request a human counsellor, with the Skill India helpline shown.
Behind the scenes, an admin dashboard gives scheme administrators a view of why families resist and where. Access is role-based: visitors, demo users, signed-in families and verified admins each see different parts of the portal.
The app is built with React and Vite, uses Firebase Authentication for email and Google sign-in, and is deployed on Vercel. It is a working prototype. The chat uses prepared, rule-based responses and the outcome data is sample data. The roadmap is an LLM grounded in verified official datasets, a real database and server-side admin roles.

Key features

  • Joint Family AI counselling:

    Joint, Parent and Learner modes in English, Hindi, Assamese, Bengali and Tamil.

  • Voice support:

    speech input and read-aloud answers, so low-literacy families can take part.

  • Verified Outcomes and NSQF paths:

    salary, placement, training centres and progression ladders per trade.

  • Family Context Explainer & ROI:

    vocational vs degree compared on cost, time and earnings, based on location and household income.

  • Human counsellor escalation:

    callback request with a reference number and the Skill India helpline (1800-123-9626).

  • Admin intelligence dashboard:

    objection reasons, district hotspots and live escalations for scheme officers.

  • Role-based access control:

    reusable ProtectedRoute, plus an admin-only page that requires a verified account on an approved email list.

  • Secure authentication:

    Firebase email/password, Google sign-in and a demo mode; passwords are never handled by the app.

  • Career exploration:

    Science, Arts, Commerce and Vocational streams, a searchable Careers section with detail pages, and job listings.

  • Production deployment:

    environment-based configuration, SPA routing rewrites and live hosting on Vercel.

Tech stack

  • React+Vite
  • Node.js
  • Firebase
Live demo Code
Where is my Bus project screenshot

Where is my Bus

Not deployed

A full-stack smart transit platform built for Assam State Transport Corporation (ASTC) city buses in Guwahati.

  • React+Vite
  • Node.js/Express
  • Google Maps

Where is my Bus

Overview

A full-stack smart transit platform built for Assam State Transport Corporation (ASTC) city buses in Guwahati, so commuters can see where their bus is.

Key features

  • Built around ASTC city bus routes in Guwahati
  • Map-based view powered by Google Maps
  • React + Vite frontend with a Node.js/Express backend

Tech stack

  • React+Vite
  • Node.js/Express
  • Google Maps
Not available online

Experience

Internships and certificates.

Research Intern, Department of Electronics & Electrical Engineering, IIT Guwahati

July 2026
Indian Institute of Technology, Guwahati

Research Intern at IIT Guwahati, benchmarking BFV and CKKS homomorphic encryption (Microsoft SEAL) to study communication overhead under different network conditions.

View certificate

Career Counselling Intern (Social Internship)

July 2025
Bishnupur Tekjuri High School

Led interactive career guidance sessions for 60+ school students in rural Assam, introducing them to new career paths and helping them navigate stream selection, parental pressure, and early goal setting.

View certificate

Participant - LabCrate workshop

July 2022
Indian Institute of Technology, Guwahati

Participated in a two-day hands-on LabCrate Workshop organized by Altanostics Labs at IIT Guwahati Research Park (July 2022), gaining early exposure to real laboratory tools and research practices while still in Class 12.

View certificate

Contact Me