Hello, I'm
Satyam
AI/ML Engineer & Agentic AI Specialist
I bridge the gap between complex ML models and production servers. From creating context-aware RAG pipelines to orchestrating multi-agent AI campaigns, I turn data into scalable, practical software.Let's build systems that actually work.
About Me
Who I Am
I build AI systems that ship. Three internships taught me what the gap between a working demo and a real deployment actually looks like.
At Asvix I built the RAG pipeline behind DigiLab, an educational chatbot running 500+ daily queries. At Cloudily Scripts I cut query latency from 8.2s to 1.7s on a live PDF RAG system. At IPtechhub I automated deployments that went from 2 hours to 15 minutes.
I also co-authored research on hybrid AI syntax detection — AST parsing plus a Gradient Boosting classifier across five languages. Preparing it for IEEE submission now.
My approach: make it simple, make it work, then make it better. Whether it's a multi-agent campaign system or a FastAPI backend, I build things people can actually use.
Education
B.Tech AI & ML
United Institute of Technology
Graduating 2026
Languages
Python, SQL
English, Hindi
Fluent in all
Focus
LLM Systems
RAG Pipelines
Backend Architecture
Approach
Full Lifecycle
Scalable Solutions
Production Ready
Work History
My Journey
AI Developer Intern
- Built the embedding pipeline for DigiLab, an AI chatbot using LangChain + FAISS + Neo4j hybrid RAG — handling 500+ daily queries at 99.2% uptime.
- Reduced hallucination rate from 18% to 11% through context-aware response modules and iterative prompt refinement.
- Improved medical query relevance by 23% using optimised FAISS vector retrieval and BM25 hybrid search.
- Integrated LLM evaluation metrics (faithfulness, context recall) for data-driven prompt iteration.
AI Chatbot Development Intern
- Built a production RAG pipeline (FAISS IVF128 indexing, cross-encoder reranking, BM25 dense retrieval) that processed 100+ page PDFs at 91% accuracy and cut support tickets 35%.
- Cut query latency 79% (8.2s → 1.7s) with parallel embedding and semantic chunking; Dockerized the stack and reduced image size 60% (2.1 GB → 840 MB).
Cloud Engineering Intern
- Deployed containerized ML inference services on AWS EC2 with auto-scaling; handled 500+ daily requests at 99.5% uptime and cut infrastructure costs 32%.
- Automated CI/CD via GitHub Actions; deployment time dropped 87.5% (2 hrs → 15 min); cold-start improved from 45s to 8s.
Portfolio
Things I've Built
Here are some projects I'm proud of. Each one taught me something new about building practical AI solutions.
Multi-Agent Campaign Creator
Advanced multi-agent AI system for creating marketing campaigns. LangGraph manages the state machine between agents. CrewAI handles each agent's execution.
RAG Pipeline
4-agent pipeline: Retriever, Guardrail Agent, Generator, Evaluator. RAGAS evaluation harness measuring faithfulness, context recall, context precision. OpenTelemetry observability. Retry logic via tenacity. Dockerized. GitHub Actions CI.
HybridAI Syntax Error Detection
Dual-mode syntax error detection across 5 languages. AST rule-based analysis runs first; a Gradient Boosting classifier catches the rest. 94.18% accuracy, Cohen's κ 0.79, ~1ms median inference. FastAPI server, Streamlit UI, CLI. Graceful degradation when the ML model can't load.
Customer Churn MLOps Pipeline
End-to-end ML pipeline with a 3-model benchmark (LR 0.849, RF 0.861, XGBoost 0.868 ROC AUC), served via a FastAPI REST API and Dockerized for reproducible deployment. Batch prediction endpoint with data-drift monitoring.
LLM Judge Evaluation
Research-backed LLM evaluation framework — pairwise and pointwise scoring with position bias, verbosity bias, and self-enhancement detection. Based on MT-Bench, G-Eval, and Prometheus. FastAPI REST API, Streamlit UI, and Python SDK.
Tech Stack
What I Work With
These are the tools and technologies I use to bring ideas to life. Always learning, always experimenting.
Languages
AI & LLM
Backend
Databases
ML & Data
DevOps & Cloud
Skill Constellation
Daily Tools
Get In Touch
Let's Connect
Got an interesting project or just want to chat about AI? I'm always up for a conversation. Drop me a message!
Primary Contact
Open to freelance work and full-time roles