Initializing AI...
đ Hello, World! I'm
AI Developer & Data Scientist specializing in LLMs, RAG, Agentic AI, and GenAI systems. Turning complex data into intelligent, production-ready solutions.
Years Exp
Projects
Technologies
Certifications
class AdityaGupta:
role = "AI Developer"
location = "Pune, Maharashtra"
education = "B.Tech CSE @ Parul Univ"
cgpa = 8.46
skills = [
"LLMs & RAG",
"Agentic AI",
"LangChain / LlamaIndex",
"GenAI & Prompt Eng.",
"AWS & MLOps",
]
def solve(self, problem):
return self.build_ai(problem) â¨
// what i do
Building production-grade LLM applications with RAG, fine-tuning, and Agentic AI using LangChain, LlamaIndex & CrewAI.
Designing and deploying ML models for NLP, Computer Vision, and predictive analytics using TensorFlow & PyTorch.
Deploying scalable AI pipelines on AWS with Docker, CI/CD, MLflow, and DVC for end-to-end ML lifecycle management.
Extracting insights from complex datasets using statistical analysis, visualization, and advanced data engineering techniques.
// about me
I'm a passionate AI Developer & Data Scientist currently working at Edunet Foundation as a Subject Matter Expert, designing and delivering AI curriculum to the next generation of developers.
With 1+ year of experience in designing intelligent retrieval systems, fine-tuning LLMs, and deploying scalable AI pipelines on AWS. My work spans from NLP and GenAI to MLOps and cloud infrastructure.
Previously worked at Metapercept Technology and interned at Vision Infotech and University of West London (Research), where I built AI-powered products used by 500+ students daily and published research on Stable Diffusion.
// tech stack
// my work
Multi-Domain AI Chatbot using LangChain & RAG architecture. Integrated ChromaDB for knowledge retrieval with sub-2-second latency across 1,000+ knowledge base entries.
Sub-2s latency
1,000+ KB entries
Multi-domain
RAG Architecture
Platform connecting to Instagram accounts to auto-detect & remove abusive comments in real-time. Fine-tuned BERT on custom Hinglish + English dataset achieving 91%+ accuracy.
91%+ Accuracy
Real-time filter
Instagram API
Hinglish + English
AI-driven platform using Hybrid RAG (Vector DB + Knowledge Graph) enabling intelligent retrieval across 1,000+ documents, reducing query latency by ~35%.
35% less latency
1,000+ docs
Knowledge Graph
AWS OpenSearch
RAG-based system generating curriculum-aligned question papers for 500+ NEET & JEE students daily. Reduced paper delivery time from 2 hours to under 30 seconds.
500+ students/day
2hr â 30sec
30% relevance boost
NEET & JEE
Custom Stable Diffusion techniques for text-to-image generation achieving 20â40% improvement in FID score over baseline. Published research papers & technical reports.
20-40% FID boost
Text-to-Image
Published Research
Univ. West London
Asynchronous file processing & OpenSearch indexing pipeline improving document ingestion throughput by 2x and reducing processing bottlenecks at scale.
2x throughput
Async processing
AWS Lambda
Dockerized
// my journey
India
Remote
Pune, Maharashtra
Alwar, Rajasthan
Vadodara, Gujarat
Microsoft Certified
Jul 2024NPTEL
Jun 2024DeepLearning.AI / Coursera
Sep 2023Microsoft Certified
Feb 2023NPTEL
2023Udemy â Andrei Neagoie
2023// get in touch
Have an AI project in mind? Let's build something amazing.