Open to opportunities

PRERNA MISHRA

@prernamishra4

Message

AI/ML engineer building production-grade GenAI, NLP, and agentic systems.

India

Message

What I'm looking for

I’m seeking a junior AI/ML role where I can contribute to real-world LLM and agentic AI systems while continuing to grow—building production-grade GenAI/NLP components like RAG, REST APIs, and reliable deployments.

I’m a motivated AI/ML Engineer pursuing B.Sc. (Hons) Computer Science at the University of Delhi, and I focus on building production-grade GenAI, NLP, and machine learning systems. I’ve developed three end-to-end AI applications featuring multi-agent orchestration, RAG pipelines, REST APIs, and vector databases—work that reflects the practical engineering depth expected in junior roles.

In my projects, I architect production-style RAG systems that connect document ingestion to vector indexing and LLM-based generation. I built a FastAPI-based Interview Preparation Assistant with a multi-endpoint workflow (question generation, answer evaluation, session history) and CI via GitHub Actions, using Sentence Transformers with FAISS and Gemini for retrieval and response generation.

I also engineer agentic experiences and deploy full ML lifecycle apps with reliability in mind. My Multi-Agent Research Assistant demonstrates Planner → Search → Summarizer → Writer orchestration with live web retrieval via Tavily and structured reporting, while my Student Exam Performance Predictor covers end-to-end ML (EDA, feature engineering, benchmarking, model serialization with Joblib) and deployment with fast API responses on Render.

Experience

Work history, roles, and key accomplishments

Multi-Agent Research Assistant

GitHub

Engineered a 4-agent research platform (Planner→Search→Summarizer→Writer) using FastAPI and Google Gemini with live web retrieval via Tavily. Built a document-based Q&A RAG pipeline using ChromaDB and HuggingFace embeddings, delivered 10+ Swagger/OpenAPI REST endpoints with SQLAlchemy-backed session history, and containerised the app with Docker.

fastAPI Google Gemini Tavily RAG ChromaDB SQLite Docker OpenAPI SQLAlchemy

Interview Preparation Assistant

GitHub

Architected a production-style RAG pipeline (PDF→FAISS vector indexing→Gemini) to generate interview questions and evaluate answers using semantic similarity. Delivered 6 FastAPI REST endpoints with Swagger/OpenAPI and GitHub Actions CI, achieving sub-100ms retrieval and >90% relevance on manual test cases.

fastAPI RAG Faiss Gemini Sentence Transformers REST APIs OpenAPI GitHub Actions CI CD

Student Exam Performance Predictor

GitHub

Developed and deployed an ML web app to predict student Math scores, covering EDA, feature engineering, benchmarking 5+ algorithms, and model serialization. Selected CatBoost as the production model and deployed the service on Render with <500ms average API response time.