॥ AI Engineer · RAG & LLM Systems · VIT 2026 ॥
NILESH
PANDEY

I forge machines that think, systems that learn,
and models that understand — at scale.

View My Work Seek Alliance
NEECHE JAIYE
॥ 01 — Parichay

AI that
scales

✦

B.Tech student at Vellore Institute of Technology, obsessed with building AI systems that solve real problems at massive scale. From deploying RAG pipelines handling 1,000+ daily queries to fine-tuning LLMs for medical diagnosis with 97% accuracy — I bridge research and production.


Ex Intern at Prxis Tech Solutions, engineered scalable AI infrastructure with Llama 3, FAISS, and cloud-native systems. Grand Finalist & 1st Place Healthcare at IIC 2024.

PROFILE · PATRIKA
❯ cat profile.json
{
  "naam": "Nilesh Pandey",
  "karya": "AI/ML Engineer",
  "vidyalaya": "VIT 2026",
  "cgpa": 7.70,
  "vishesh": [
    "LLM Fine-tuning",
    "RAG Systems",
    "Computer Vision",
    "NLP at Scale"
  ],
  "sthiti": "Open to opportunities"
}
❯
॥ 02 — Karya — Selected Work

Built to matter

// NLP · Distributed Systems
Twitter Mental Health Crisis Detection
Distributed NLP pipeline analyzing 42M+ tweets in real-time using VADER sentiment analysis, parallel batch processing, and Plotly dashboards for crisis detection at scale.
42M+Tweets Analyzed
Selenium+ NLTK + VADER
// LLM · RAG · Medical AI
AI-Based Symptom Detector
Fine-tuned Mistral 7B Instruct with FAISS vector search and Sentence Transformers. Cloud-deployed with modular NLP pipeline and API-based retrieval.
97%Accuracy
1000+Test Cases
// CV · Deep Learning · Security
Facial Authentication + Anti-Spoofing
CNN-based facial auth replacing CAPTCHA/passwords. TensorFlow + OpenCV + MediaPipe with SSL/TLS security and real-time anti-spoofing detection.
98%+Recognition Acc.
85%↓Spoof Attacks
// LLM · RAG · Cloud Infra
Multi-Agentic RAG System @ Prxis Tech — Production Grade
Engineered and deployed a scalable RAG system using Meta Llama 3 8B, FAISS vector database, and Sentence Transformers. Reduced inference latency by 40%, preprocessing time by 30%, enabling cloud deployment for 10K+ daily queries.
40%Latency ↓
30%Preprocessing ↓
1K+Daily Queries
Llama 38B Model
॥ 03 — Shakti — Technical Arsenal

Tools of the craft

LLMs SLMs LangChain FAISS RAG Systems PyTorch TensorFlow Keras FastAPI Flask Python C++ SQL MongoDB PostgreSQL AWS Google Cloud OCI HuggingFace OpenCV Pandas NumPy Scikit-learn MLOps
॥ 04 — Gaurav — Recognition

Forged under pressure

IIC 2024 · International Innovation Challenge
Grand Finalist + 1st Place
Healthcare Category

From 764 competing teams, secured Top 10 in the grand finals. In 48 hours, designed and deployed a full AI-driven medical assistant — fine-tuned Mistral 7B in a RAG system for real-time symptom detection and diagnosis.

764Competing Teams
Top 10Grand Finals
48hBuild Time
#1Healthcare Track
॥ 05 — Praman — Certifications

Verified expertise

✦
Oracle
Oracle
Data Science Professional
OCI Certified · Verified Badge
Verify Credential →
IBM
IBM
Generative AI
IBM CEP · e8aa2dc4124b4030
Verify Credential →
CEH
EC-Council
Certified Ethical Hacker v12
Cert No · ECC9680215374
Verify Credential →
Udemy
Udemy
Data Science: Transformers & NLP
UC-227bf3a5
Verify Credential →
Coursera
Coursera · Google
Bits and Bytes of Networking
Verified · 34c614dd84a71a3b
Verify Credential →
॥ 06 — Sankhya — GitHub

The code speaks

43Repositories
AI · ML · DSPrimary Domain
VITBatch · 2026
GitHub Stats
GitHub Streak
View GitHub Profile →
॥ 07 — Sampark — Contact
LET'S
CONNECT

Open to full-time roles, internships & high-impact collaborations in AI & ML.

Email
pandeynilesh325a
@gmail.com
OPEN →
GitHub
nilesh03pandey
OPEN →
LinkedIn
nilesh-pandey-
ai-sec
OPEN →
WhatsApp
+91 902****481
OPEN →
Nilesh Pandey
AI · ML · LLM
VIT 2026
© 2025
Nilesh Pandey // 2025
AI · ML · LLM