Summary
Overview
Work History
Education
Skills
Affiliations
Timeline
Generic

Sweety Junnarkar

Pune

Summary

Dynamic data science professional with a proven track record at SRBC Co LLP, specializing in NLP and Generative AI solutions. Expert in deploying machine learning models and enhancing user experiences. Adept at collaborating with clients to deliver impactful, data-driven insights, leveraging skills in Python and effective communication to drive results.

Overview

5
5
years of professional experience

Work History

Associate Consultant(Data Science)

SRBC Co LLP (EY)
Pune
01.2023 - 08.2024
  • Conceptualized and designed robust NLP solutions and LLM agents focused on user experience, interactivity, and functionality.
  • Work with clients, understand the problems, and find data-driven solutions. Built and deployed APIs using Flask, FastAPI, and Docker for serving ML, DL, and Generative AI models.
  • Preprocessed and curated text data through cleaning, tokenization, and data augmentation to prepare datasets for utilized tools like Langchain, Embedchain, and vector databases.
  • Developed a RAG-based chatbot to generate insights from customer reviews.
  • Developed a Text-to-SQL agent using LangChain, OpenAI LLMs, and SQL database connectors to allow users to query structured data via natural language
  • Fine-tuned BERT to detect customer intent in chatbot queries (e.g., refund request, order status, cancellation), improving automated response accuracy.
  • Developed an AutoGen-based agent to generate topic-specific blog posts for product promotion.
  • Deployed ML/DL models on AWS EC2 and SageMaker using Docker, CI/CD pipelines, and workflow orchestration tools (e.g., Airflow), with infrastructure provisioned via Terraform.
  • Hands-on experience with AWS services, including Lambda, S3, Redshift, and DynamoDB, for scalable data pipelines and model integration.

Data Engineer

3loq Labs
11.2021 - 07.2022
  • Developed a machine learning POC to analyze customer transaction patterns for personalized banking product recommendations, increasing cross-sell engagement by 18%.
  • Used Generative AI (OpenAI GPT-3.5) to auto-generate personalized email and SMS scripts for marketing campaigns based on customer segmentation and product affinity.
  • Developed an LSTM Deep Learning model which takes a sequence of transactions. Through these inputs, the model learns the most probable transaction sequences basis which it predicts the next transaction after a mentioned transaction.

CRM Executive

Imuons Web Solutions
Pune
02.2020 - 08.2020
  • Analyzed sales trends, marketing campaigns, activity reports, Dashboards, and team performance with the help of Power Bi, Excel, and python.
  • Led the team for promotion campaigns and increased sales by 20%. Analyzed increased sales data to inform business development and sales department.

Education

Warananagar -

Tatyasaheb Kore Institute Of Engineering & Technology
Warananagar
05.2017

Skills

  • Programming Languages: Python
  • Database: PostgreSQL
  • Machine Learning & Deep Learning Frameworks: Logistic Regression, OpenAI, Scikit-learn, Pandas, XGBoost, Hugging Face Transformers,Numpy, Linear Regression, Decision Tree, Random Forest
  • NLP (Natural Language Processing) Tools: NLTK, SpaCy, Gensim
  • Large Language Models: Transformer Models, GPT-3/4, Gemini, Llamma2
  • Data Visualization: PowerBI, Plotly, Matplotlib
  • Cloud Computing : AWS
  • Generative AI Framework: Langchain, VectorDB,RAG, Autogen
  • Devops : Git, MLflow, DVC, Docker, CI/CD
  • ETL Tool: Alteryx

Affiliations

  • Developed an AI-based summarization tool specifically designed for financial documents like quarterly reports, balance sheets, and meeting minutes using Large Language Models (LLMs) .
  • Utilized OpenAI’s GPT-3/4 and LangChain for generating concise and accurate summaries while preserving the key financial insights and compliance information.
  • Integrated Retrieval-Augmented Generation (RAG) techniques to enhance document context retrieval, ensuring that the summaries captured all relevant data points from large, unstructured documents.
  • Built a custom data preprocessing pipeline using Python to handle various formats (PDF, DOCX, images) and structured the input for efficient processing by LLMs
  • Deployed the system with a Flask and React-based UI for seamless user interaction, allowing financial analysts to input documents and retrieve summaries instantly.
  • Deployed the system with a FastAPI for seamless user interaction, allowing financial analysts to input documents and retrieve summaries instantly.
  • Reduced the time spent on manual document review by 50%, increasing the efficiency of financial analysis and reporting workflows.
  • Technologies: Python, GPT-3/4, LangChain, RAG, FastAPI, VectorDB, and financial data analysis

Timeline

Associate Consultant(Data Science)

SRBC Co LLP (EY)
01.2023 - 08.2024

Data Engineer

3loq Labs
11.2021 - 07.2022

CRM Executive

Imuons Web Solutions
02.2020 - 08.2020

Warananagar -

Tatyasaheb Kore Institute Of Engineering & Technology

Similar Profiles

CREATE PROFILE
Sweety Junnarkar