Skip to content

DevJin27/DevJin27

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 

Repository files navigation

DevJin27

Hi — I'm Dev Jindal (DevJin27).
I'm passionate about building intelligent, reliable systems and learning continuously — from algorithms and systems to ML and full‑stack apps.

  • 💻 Focus: Back-end & full-stack development, APIs, systems design
  • 🤖 Interests: AI, fine-tuning models, applied ML, developer tooling
  • 🌱 Always learning: system design, React/Next.js, model fine-tuning
  • 📫 Reach me: [email protected] | https://github.com/DevJin27

⚙️ Core Tech & Tools

I work across the full stack and with ML tooling:

  • Languages: Python, JavaScript (ES6+), SQL
  • Back-end: FastAPI, Flask, REST, WebSocket, Docker
  • Front-end: React, Next.js, HTML/CSS
  • ML & AI: Model fine-tuning, prompt engineering, evaluation pipelines
  • Datastores: MySQL, PostgreSQL, Redis
  • Dev & Deploy: Git, GitHub, Docker, CI/CD

🔭 Currently Learning & Working On

  • Deepening knowledge of model fine-tuning and evaluation pipelines.
  • Building RISE — a DSA Mentor AI app (see project details below).
  • Improving React/Next.js front-end patterns and full-stack integration.

Projects

1) Shelf Stack

  • Tech: Python, MySQL, HTML/CSS, JS
  • Status: Completed — production-ready MVP
  • Overview: Shelf Stack is an inventory and retail management system tailored for small bookstores and shops. It combines staff management, role-based access control, inventory tracking, and sales analytics into one lightweight system.
  • Key features:
    • User authentication and role-based access (Admin, Manager, Cashier)
    • Inventory management: add/edit books, categories, stock counts
    • Sales & POS: create sales, discounts, returns, and generate receipts
    • Sales tracking & reports: daily/monthly sales summaries, bestsellers, low-stock alerts
    • Reporting: exportable CSV reports, simple dashboard metrics
  • Architecture:
    • Flask (or FastAPI) backend with REST APIs
    • MySQL for persistent storage
    • Modular service structure for inventory, sales, and users
  • Deployment & run notes:
    • Designed to run in a container or small VM; easy to deploy with Docker
    • Can integrate barcode scanners and receipt printers for POS use

2) Project_map

  • Tech: FastAPI, React (WebSockets), map visualization libraries, PostgreSQL/MySQL
  • Status: Completed — realtime telemetry tracking system
  • Overview: Project_map captures, processes, and visualizes real‑time telemetry from drones (or similar devices). It’s built to ingest frequent telemetry, provide live location & status updates on a map, and store historical telemetry for replay and analytics.
  • Key features:
    • Real-time ingestion via REST + WebSocket streaming for low-latency updates
    • Interactive frontend map showing live locations, flight paths, telemetry overlays
    • Telemetry pipeline with storage for historical playback and diagnostics
    • Alerts for geofence breaches, low battery, or anomalous telemetry
    • Authentication for device operators and admin dashboards
  • Architecture:
    • FastAPI handles telemetry ingestion, WebSocket streams, and REST endpoints
    • Front-end (React) subscribes to telemetry channels for live map updates
    • Optional message broker (Redis/Redis Pub/Sub or lightweight queue) for scaling real-time events
  • How to run (example)
    # backend
    uvicorn app.main:app --reload
    
    # frontend
    cd web && npm install && npm run dev
  • Notes: Built with extensibility in mind — can add ML-based anomaly detection, geospatial indexing, or integrate with cloud streaming services.

3) RISE — DSA Mentor (in development)

  • Tech stack: FastAPI, model fine-tuning (transformer-based), React, Next.js, Python
  • Status: Active development (I'm currently working on this)
  • Vision: RISE is a DSA-focused AI mentor that helps learners prepare for algorithmic interviews and improves problem-solving skills using personalized coaching and automated feedback.
  • Core capabilities:
    • Personalized learning paths and problem recommendations based on skill gaps
    • Problem explanation generation and step-by-step hints
    • Automated code evaluation and constructive feedback (style, complexity, correctness)
    • Adaptive difficulty and progress tracking
    • Conversational mentor UI powered by a fine-tuned model (for DSA pedagogy and scaffolding)
  • Technical details:
    • FastAPI provides the API backend, evaluation sandbox, and model serving endpoints
    • Model fine-tuning pipeline for specialized DSA tutoring responses and scoring
    • React + Next.js for an interactive UI (SSR for landing and SEO; CSR for the mentor/chat experience)
    • Secure code execution sandbox for running user solutions (isolated, resource-limited)
    • Analytics and telemetry to track learning progress and model performance
  • Current focus:
    • Building a robust evaluation and sandboxing system for running submissions safely
    • Fine-tuning model(s) with curated DSA dialogues and high-quality feedback examples
    • Implementing the mentor chat experience and personalized plan generator
  • Example dev steps
    # backend dev
    uvicorn rise_api.main:app --reload
    
    # frontend dev
    cd client && npm install && npm run dev
  • Goal: early private alpha → public beta with instructor/dev tools and integrations (GitHub, LeetCode-style problem import, etc.)

How to collaborate / contribute

  • Want to try RISE or contribute to Project_map/Shelf Stack? Reach out at [email protected] with a short note about how you'd like to help.
  • If you find issues or want features, open an issue or PR in the repo (linked project repos will have contributing guidelines).

Contact & Social


Thank you for stopping by — I enjoy working on systems that combine solid engineering with intelligent behavior. If you'd like to see code, demos, or discuss collaboration, drop me a message!

About

Config files for my GitHub profile.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors