Utkarsh Srivastava

AI Researcher

South Delhi, Delhi, India5 yrs 11 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Expert in generative AI and LLMs.
Led innovative document processing solutions.
Strong collaboration and strategy in AI projects.

Stackforce AI infers this person is a SaaS and Fintech expert specializing in AI-driven document processing solutions.

Contact

Skills

Core Skills

Large Language Models (llm)Generative AiData ScienceArtificial Intelligence (ai)

Other Skills

PandasAILangChainLlamaIndexRAGGPT-4oLLaMAMistralDeepSeekPythondockerDeep LearningAWSLLMBERTLayoutLM

About

At Newgen Software, our team is revolutionizing document analysis with cutting-edge AI. I lead innovation in table detection, document structure recognition, and generative AI for document processing. Applying state-of-the-art algorithms like LayoutLM and BERT, and LLMs like Mistral, llama3, OpenAI GPT4 we tailor transformative solutions for our clients. Collaboration and strategy are at the core of our success. We've deployed transformer-based models to enhance understanding and extract insights from complex documents. My commitment to exploring and implementing generative AI, alongside a proficiency in FastAPI and cloud platforms, drives our team’s continual growth and the elevation of our data science capabilities.

Experience

5 yrs 11 mos

Total Experience

1 yr 6 mos

Average Tenure

1 yr 5 mos

Current Experience

Staqu technologies

Senior Research Scientist

Dec 2024 – Present · 1 yr 5 mos · India · On-site

Worked with LLM tools like PandasAI, LangChain, and LlamaIndex to build chatbots for analytics, search solutions and summarization.
Developed and optimized RAG (Retrieval-Augmented Generation) pipelines for fast information retrieval.
Worked with both on-premise and cloud-based LLMs, including GPT-4o, LLaMA, Mistral, and DeepSeek.
Implemented model quantization, sharding, and fine-tuning to optimize LLM performance.
Applied prompt engineering techniques to enhance model response quality and efficiency.
Integrated structured and unstructured data into LLM workflows for better contextual understanding.
Leveraged vector databases for semantic search and knowledge retrieval.
Developed scalable search solutions using LLM-based architectures.
Automated document summarization and retrieval using customized RAG pipelines.

PandasAILangChainLlamaIndexRAGGPT-4oLLaMA+4

Newgen software

2 roles

Senior Data Scientist

Promoted

Oct 2023 – Dec 2024 · 1 yr 2 mos · Noida, Uttar Pradesh, India · On-site

Working with a team of data scientists, foster a collaborative and innovative team culture.
Apply advanced data science techniques to solve complex problems, utilise various AI/ML algorithms and deep learning models for document understanding tasks.
Working on (Patent) state-of-the-art table detection and structure recognition algorithm.
Engage with clients to understand specific needs and requirements.
Contribute to the development of the overall data science and analytics strategy.
Implement generative AI, LLMs and RAG pipelines for structure recognition and information extraction task.
Build and deploy various transformer-based models, LLM pipelines, IDP pipelines

PythondockerData ScienceGenerative AI

Data Scientist

Dec 2022 – Oct 2023 · 10 mos · Noida, Uttar Pradesh, India · On-site

Engaged in the development of an innovative intelligent document processing product for BFSI sector
Implementing use cases of LLM models e.g. RAG pipelines, for real time business applications, improving product performance, accuracy and costumer experience.
Leveraging my expertise as a data scientist, I am employing various AI/ML algorithms and deep learning transformer based models to achieve SOTA results in document classification, Information extraction and visual question answering task.
Working on various cloud platforms e.g. (aws, azure, gcp).
I adeptly utilize technologies that help in building scalable AI applications and deploy them end to end.
Leveraging GenAI capabilities into current products for enhanced performance of current services by fine tuning LLm models and using various Prompt Engineering techniques like CoT, RAG etc.

AWSLLMData ScienceArtificial Intelligence (AI)

Lumiq

Data Scientist

Jan 2021 – Dec 2022 · 1 yr 11 mos · Delhi, India

Working on building end to end product features and services and delivering solutions to major leaders of Banking sector.
Implementing models like BERT, LayoutLM, xlmRoberta to derive SOTA results on Information Extraction, Classification and Question Answering task.
Working on Natural Language Processing, quantitative analysis, predictive modeling, and data visualization to provide best data insights
Work with various AWS services like S3, EC2, Textract, SageMaker, Lambda, DynamoDB, worked on PYTHON, Docker, flask, FastApi and made various successful deployment of services on Prod.

AWSPythonData ScienceArtificial Intelligence (AI)

The intect

Data Science jr.

May 2020 – Dec 2020 · 7 mos · Delhi, India

Worked on Genetic Algorithm, Random Forest and Swarm intelligence.
worked on real time problems and live projects, explored various ML and deep learning models and their implementations using transformers and other techniques.

AWSPython

Safe security

Ethical Hacking

Jun 2018 – Aug 2018 · 2 mos · New Delhi Area, India

Worked on different aspects of cyber security and implementation of AI in improving penetration testing tasks.
Used Python as base language to build Rainbow table attacks and it's prevention, build keyloggers to monitor actions of users on company's server

AWSPython