Pranab P.

Key Skills

Software

Don't disclose

Top Subject Matter

LLM assistant for retail customer support and sales outreach

LLM-based entity/category mapping (audiences, brands, domains to IAM categories)

Anomaly classification and ontology mapping using ML and LLMs

Top Data Types

Document

Text

Top Task Types

Prompt + Response Writing (SFT)

Function Calling

Entity (NER) Classification

Freelancer Overview

AI Engineering Freelancer (LangChain/LangGraph LLM assistant; multi-agent RAG + prompt engineering). Brings 10+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose, Databricks, and MLFLOW. Education includes Bachelor of Engineering, BITS Pilani (2017). AI-training focus includes data types such as Computer Code, Programming, and Document and labeling workflows including Prompt + Response Writing (SFT), Function Calling, and Entity (NER) Classification.

Labeling Experience

AI Engineering Freelancer (LangChain/LangGraph LLM assistant; multi-agent RAG + prompt engineering)

Don't disclosePrompt + Response Writing (SFT)

Built and deployed an LLM assistant workflow that handles customer pre-sales and post-sales interactions for an online retail business. Implemented a multi-agent system combining web-sourced insights with retrieval-augmented generation and dynamic prompt engineering to produce context-aware outputs. Focused on generating tailored conversational responses and personalized outreach emails rather than manual labeling. • Defined LLM assistant behavior and multi-agent orchestration for sales/support conversations • Used RAG and prompt engineering to tailor tone, messaging, and value propositions • Generated personalized outreach emails based on dynamically gathered pain points • Integrated the assistant into WhatsApp and a web application for end-user interaction

2025 - Present

Senior Data Scientist (LLM-powered mapping pipeline and MLOps)

DocumentFunction Calling

Designed and led development of an LLM-powered market segment to audience mapping pipeline. The work involved mapping audiences, brands, and domains to IAM categories using LLM outputs, including deployment on internal GPU servers and Databricks. While not manual annotation, it constitutes structured AI training/mapping to produce categorical labels for downstream use. • Scoped and developed LLM-powered mapping of entities to IAM categories • Built and operationalized MLOps for an LLM-powered mapping workflow using Databricks, MLflow, and Jenkins • Implemented CI/CD and modular best practices for maintainable ML code • Led mentorship and reviews for junior data scientists involved in the pipeline

2023 - 2025

Data Scientist II (classification models for ontology/entity labeling from logs)

TextEntity (NER) Classification

Built machine learning models to detect and classify anomalies and map them to relevant ontology classes and objects. Developed an LLM classifier fine-tuned on system logs combined with numerical features to generate categorical decisions. Delivered supporting production-ready ML pipelines for training and inference. • Developed multi-class ensemble classification for anomaly-to-ontology mapping • Trained an LLM classifier using system logs plus structured numerical features • Implemented end-to-end ML/AI pipeline with Python, Airflow, and feature engineering • Created CI/CD ML pipeline repository for proof-of-concept fine-tuning and iteration

2021 - 2023

Data Analyst (AI-powered essay assessment and NLP metrics)

Text

Performed NLP-driven analysis to generate evaluation metrics and insights from a corpus of abstractive essays for downstream assessment. Implemented text preprocessing and semantic analysis using common NLP libraries and embedding/NER techniques. Developed an AI-powered program that generates assessment reports from handwritten essays and exposed it via a deployed REST API. • Generated evaluation metrics and assessment strategies from essay text • Implemented text preprocessing (NLTK, spaCy), embeddings (GloVe), and NER models • Built an AI assessment report generator for handwritten essays and deployed a REST API • Developed end-to-end pipelines for data collection, cleaning, training, and inference for deployed features

2019 - 2020

Education

Work History