Chandan Singh

Chandan Singh

B.Tech CSE student · AI/ML Engineer · Computer Vision · NLP · Full-Stack AI Systems

Noida, Uttar Pradesh, India

@chandan_singh

0 followers

Looking for internships

🎓 B.Tech, Computer Science & Engineering at Maharishi University of Information Technology · Graduating 2027

About

B.Tech CSE student specializing in AI/ML engineering with hands-on experience building deployable computer vision, NLP, and full-stack AI systems. Has shipped multiple end-to-end projects using YOLOv8, LSTM, XGBoost, SegFormer/DeepLabV3+, and LLM-based agentic pipelines.

Education

Skills

2 to 5 years 2 to 5 years Several years of experience.
1 to 2 years 1 to 2 years A year or two of steady use.
3 to 6 months 3 to 6 months A few months of practice with it.

Tools / apps / platforms

1 to 2 years 1 to 2 years A year or two of steady use.
Vite Hugging Face Hub
6 to 12 months 6 to 12 months Most of a year behind them.
Git PostgreSQL
<3 months <3 months Just getting started - under three months. This is also what shows when a level has not been set.
Docker

Projects

  • NikaAI
    YOLOv8s, FastAPI, Celery, Vercel

    Industrial defect detection and quality intelligence platform for the AMD AI Hackathon. Fine-tuned YOLOv8s for 17 defect classes, added uncertainty quantification and a Gemma LLM reasoning layer, and published the detector on Hugging Face Hub.

  • HR-RAG-Assistant
    LangChain, FAISS, Jina Embeddings, Groq, Streamlit

    Retrieval-augmented HR policy Q&A system with a LangChain agent, Streamlit chat UI, and persisted FAISS index.

  • MedVoice-AI
    React, FastAPI, OpenAI, Sarvam, Hugging Face, Groq

    Agentic clinical documentation platform for visit records, audio transcription, EMR generation, consent logging, and PDF export.

  • HABS
    FastAPI, React, Random Forest, SHAP, Alembic

    Healthcare appointment booking system with async SQLAlchemy backend, no-show prediction model, and explainability.

  • TRINETRA
    XGBoost, Twilio, Google Safe Browsing

    AI scam intelligence engine for real-time fraud detection across URLs, SMS, and WhatsApp screenshots.

  • SIGNSPEAK
    LSTM, MediaPipe, React, Vite, Flask

    Word-level sign language recognition system using 30-frame LSTM sequences over MediaPipe hand landmarks.

  • Falcon Offroad Semantic Segmentation
    DeepLabV3+, SegFormer, Albumentations

    Offroad semantic segmentation model for HackWithMumbai 2.0 on Duality AI's Falcon desert digital-twin dataset.

Courses & certifications

  • AI/ML for Geodata Analytics · Indian Institute of Remote Sensing (IIRS), ISRO, Dept. of Space · 2026

🏆 Achievements & awards

  • AMD AI Hackathon

    Designed and deployed NikaAI, a production-oriented computer vision platform, in a hackathon track sponsored with a $20,000+ overall prize pool.

  • HackWithMumbai 2.0

    Built the Falcon offroad semantic segmentation model on Duality AI simulation data as part of the BWI-PS track.

  • Self-directed model publishing

    Released a fine-tuned defect-detection model on Hugging Face Hub for public reuse and reproducibility.

Explore the Broxer community

🤖
Online · instant AI help
Broxer