Abhinav Ennazhiyil

Data Scientist

Profile

Data scientist with a Bachelor's degree in Computer Science and 3 years of work experience. Proven track record of building and deploying ML and LLM-based solutions that drive business impact across marketing, sales, and finance. Developed a high-performing CLTV model from scratch and designed AI systems that generate personalized marketing emails and financial reports using Retrieval-Augmented Generation (RAG) frameworks. Skilled in working cross-functionally with stakeholders to translate complex business challenges into scalable data products. Strong technical foundation in machine learning, LangChain, AWS Bedrock, and data experimentation, with a focus on delivering actionable insights and measurable outcomes.

Skills

LangChain PydanticAI RAG AWS Bedrock Machine Learning (Regression, Classification, Clustering) Statistical Analysis Python Excel SQL Project Management

Professional Experience

Data Scientist

Jun 2022 – present

C2FO | Noida, Delhi

AI Email Generator

  • Built C2FO's first production-grade GenAI application, increasing average response rates from ~1% to ~5% across Europe and the U.S.
  • LLM-driven marketing email generator using LangChain and AWS Bedrock, integrated into Marketo.
  • Email personalization using real-time Internal data, conversation history, and external data like market trends, press releases, etc.
  • Implemented quality assurance using a custom validation system (DeepEval).

AI Financial Reports

  • Developed a financial report generator to support the marketing team with customer's working capital insights, reducing research effort from ~3 days to ~2 min.
  • Used a RAG pipeline with LangChain and AWS Bedrock to extract and synthesize key trends from SEC 10-K/10-Q filings.
  • Integrated the reports into Salesforce for on-demand sales enablement.

Customer Lifetime Value Model

  • Engineered a custom Customer Lifetime Value model using XGBoost with ~70% recall.
  • Defined customer lifetime value segments using a RFM matrix for a period of 1 year from first transaction.
  • Tested the performance using a A/B Testing framework in collaboration with the campaign management team.

Trigger based sales prioritization

  • Developed a custom spike‑detection logic for significant jumps in the customer's accounts payable and day sales outstanding.
  • Integrated into the sales‑targeting system to prioritize high‑impact accounts.

Projects

The LLM Podcast

An LLM driven sports news website

  • Curated real-time sports news by scraping articles from multiple online sources, leveraging LLM agents to extract and synthesize relevant insights.
  • Implemented unsupervised clustering algorithms (cosine similarity with gemini embeddings) to group similar articles, followed by LLM-based summarization.
  • Developed an automated podcast engine where two distinct LLM personas engage in dynamic, human-like dialogue.

Education

Bachelor of Technology in CSE

2018 – 2022

IIIT Delhi

8.4 CGPA