Umang Gupta

AI Researcher

4 yrs 11 mos experience

AI ML PractitionerAI Enabled

Key Highlights

Expert in AI-driven video generation and multimodal systems.
Proven track record in scalable distributed systems at Amazon.
Strong foundation in Electronics and Instrumentation Engineering.

Stackforce AI infers this person is a SaaS-focused AI Engineer with expertise in machine learning and scalable systems.

Contact

Skills

Core Skills

GenaiDistributed SystemsMachine LearningDeep LearningData ScienceSoftware DesignRetrieval-augmented Generation (rag)JavascriptAwsNode.js

Other Skills

Context EngineeringPython (Programming Language)Large Language Models (LLM)Hugging FacePyTorchAzure AI FoundryAWS SageMakerRAGLLMLarge Language Model Operations (LLMOps)LangChainJavaReact.jsAmazon Web Services (AWS)SQL

About

I am an AI engineer, currently working at The D. E. Shaw Group to help build AI tools to support risk analysis. At Pictory, I helped leverage AI to democratize video creation. I contributed to improving processes from script generation to ML-powered visual search, selection, and video design. My work focused on enhancing storytelling and video creation using deep learning and multimodal AI models, leveraging PyTorch, Transformer-based models, and scalable ML pipelines. Previously at Amazon, I developed scalable distributed systems and inventory management tools, gaining expertise in system design and implementation. A graduate of BITS Pilani in Electronics and Instrumentation Engineering, I enjoy solving complex problems using data-driven approaches powered by machine learning.

Experience

4 yrs 11 mos

Total Experience

1 yr 6 mos

Average Tenure

2 mos

Current Experience

The d. e. shaw group

Senior Member, Tech (Gen AI Specialist)

Mar 2026 – Present · 2 mos · Hyderabad, Telangana, India · Hybrid

GenAIRetrieval-Augmented Generation (RAG)Context EngineeringDistributed Systems

Pictory

2 roles

SDE II (AI)

Promoted

Mar 2024 – Mar 2026 · 2 yrs

Worked on AI-powered video generation systems, building end-to-end pipelines across LLMs, multimodal retrieval, and search systems. My work spanned agentic workflows (LangGraph), RAG architectures, and LLM fine-tuning (DPO), along with embedding models and vector databases (Pinecone). Focused on improving search relevance, content generation, and personalization at scale using tools like PyTorch

Machine LearningPython (Programming Language)Large Language Models (LLM)Hugging FaceDeep LearningPyTorch+2

Software Engineer I (AI)

Mar 2023 – Mar 2024 · 1 yr

Built RAG systems for media reranking, reducing incorrect selections by 71%.
Led zero-downtime migration of 100M+ customer records while managing payment processing, subscription renewals, and quota systems.

Python (Programming Language)Software DesignData ScienceRetrieval-Augmented Generation (RAG)RAGLLM+3

Amazon

Software Development Engineer

Dec 2021 – Mar 2023 · 1 yr 3 mos · Bengaluru, Karnataka, India

Designed an inventory migration service processing 5M+ inventory records in under 5 minutes and built a unified source of truth by reconciling 1B+ records across three systems.
Built ReactJS micro front-end components for warehouse pick operations deployed across Amazon's fulfillment network

JavaScriptJavaReact.jsAmazon Web Services (AWS)Node.jsSQL+2

Doubtnut

Software Developer

Jun 2021 – Dec 2021 · 6 mos · Gurugram, Haryana, India

Built real-time quiz platform using Node.js, Redis, MongoDB and MySQL serving 100K+ daily users with sub-50ms latency.
Created automated detection tool for improperly indexed content and optimized SQL queries, improving API performance by 60%.

JavaScriptPython (Programming Language)JavaReact.jsAmazon Web Services (AWS)Node.js+4