AI Engineer at Prodigal

Naman Chhaparia

Artificial Intelligence Engineer

I build RAG systems, generative AI, and computer-vision applications that turn messy real-world data into products people actually use.

Download CV Get in touch
Portrait of Naman Chhaparia

About

I am a fourth-year B.Tech (Artificial Intelligence) student at SVKM's NMIMS - MPSTME, and an AI Engineer at Prodigal. Over the past three years I have shipped work across sentiment analysis, retrieval-augmented generation, generative AI, and OpenCV-based computer vision.

I have been a national hackathon and ideathon finalist, ranking among the top 10 teams in the country, and previously served as Head of the Department of Artificial Intelligence at the Google Developer Students Club, MPSTME.

3+ years

AI Engineer

GDSC

Former Head of AI Dept.

B.Tech

Artificial Intelligence

Experience

The stack I reach for, and where I have put it to work.

AI Frameworks & Libraries

  • TensorFlow
  • Keras
  • PyTorch
  • OpenCV
  • Hugging Face Transformers
  • LangChain
  • LlamaIndex
  • Scikit-Learn
  • NLTK
  • spaCy
  • Pandas
  • NumPy

Tools & Platforms

  • MySQL
  • MongoDB
  • Firebase
  • Docker
  • Git
  • AWS
  • Jupyter
  • FastAPI
  • Postman
  • Pinecone / FAISS

Professional Experience

  1. Building and shipping AI systems in production: retrieval pipelines, LLM-driven workflows, and evaluation tooling that keeps model behavior measurable and reliable.

  2. Bridged customer requirements and engineering, deploying and tuning AI-driven solutions and translating real-world needs into working configurations.

    • Built a CNN image classifier that sorts images at source into four classes at 96.3% accuracy.
    • Cut costs by 13% by minimizing manual intervention on non-critical workloads.
    • Added a low-confidence flagging feature that routes below-threshold images to an administrator for review.
    View certificate

Projects

A few things I have built. Open a card for the detail.

An online platform that parses a candidate's resume with OCR, generates interview questions tailored to their experience, and returns AI-driven feedback plus an overall performance score to help job seekers prepare for real interviews.

OCRLLMsNLP
View on GitHub

An AI real-estate assistant that lets users find properties with natural-language queries. It pairs a multilingual chatbot with Azure voice interaction, vector-based property search, and the Gemini API for recommendations, plus Twilio OTP for verification.

Vector searchGemini APIAzure VoiceTwilio
View on GitHub

A price-forecasting model built on the Prophet time-series algorithm, using retail scan data to predict commodity prices for agriculture, retail, and supply-chain decisions. It captures trend, seasonality, and holiday effects, reaching a Mean Absolute Percentage Error of 0.1% to 0.3%.

ProphetTime seriesForecasting
View on GitHub

More projects, including gesture recognition with AR and predictive maintenance, live on GitHub.

Get in touch

Let's build something.

LinkedIn GitHub