Default profile banner
DJ

Dibyanshu Jaiswal

@dibyanshujaiswal

Data Scientist at Fractal

Bangalore, Karnataka

linkedin.com/in/dibyanshu-jaiswal-198689177

FractalVellore Institute Of Technology

Dibyanshu Jaiswal is an experienced Data Scientist and NLP Engineer specializing in advanced machine learning and large language models. He has a proven track record developing complex solutions, including agentic platforms and multimodal chatbots, at companies like Fractal and Barclays. His technical expertise spans Python, Azure, AWS, and various deep learning frameworks, enabling him to solve challenging problems in NLP, Computer Vision, and RAG implementation.

Experience

Data Scientist

Fractal

•Apr 2024 - Present•Bangalore, Karnataka

Led the development of an innovative agentic platform capable of integrating structured and unstructured agents, delivering context-specific answers through customized agent avatars with a 95% relevancy rate. Spearheaded the development of a Multimodal Chatbot for Fractal, streamlined to interpret and respond to queries involving visual data, graphs, and their insights, delivering a precision rate of 91%. Orchestrated the development of the Customer Intelligence Insights (CII) pipeline at Fractal, empowering advanced analytics to enhance customer service, with an achieved accuracy of approximately 85%.

NLP Engineer

Barclays

•Aug 2022 - Apr 2024•Pune, Maharashtra

Analyzed market trends, identified untapped opportunities, devised new strategies, configured both real-time and batch data for Barclays lending and Onboarding journeys that increased revenue by 25% within 6 months. Led the development of a robust Chatbot for Barclays, designed to handle a wide range of queries related to domain knowledge, Appdynamics details, API logs, and CR (Change Request) process requests. Achieved an impressive accuracy rate of approximately 96%, automating essential processes for increased efficiency.

NLP Engineer

Stylumia

•Aug 2021 - Jul 2022•Bangalore, Karnataka

Enhanced the semantic search engine at Stylumia by leveraging ElasticSearch and advanced NLP techniques, achieving an accuracy of 92%. Upgraded the NLU part of the search engine with 95% accuracy and also created autocomplete and autosuggestion parts of the search engine.

Education

Vellore Institute Of Technology

Bachelor of Technology

CSE

Jan 2018 - Jan 2022•Grade: CGPA: 8.32

Skills

Java
Python
SAS
PCCM
PCSM
Azure Databricks
Microsoft Azure
AWS Bedrock
Fastapi
Flask
Django
Pandas
Tensorflow
Keras
Pyspark
Rasa
Langchain
Llamaindex
Langraph
CrewAI
Mysql
Sql Plus
Elasticsearch
Teradata
Pinecone
Chromadb
Redis
Machine Learning
Deep Learning
Natural Language Processing
Computer Vision
Prediction Modeling
Large Language Model
Retrieval-Augmented Generation
Automatic Speech Recognition
Multi Agent System
LLM
RAG
Librosa
Spectrogram
Reinforcement Learning