Anurag Singh

Data Scientist

Noida, Uttar Pradesh, India1 yr 10 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Expert in developing intelligent systems and NLP solutions.
Proven track record in deploying scalable machine learning models.
Strong background in computer vision and data extraction techniques.

Stackforce AI infers this person is a Data Scientist specializing in AI-driven solutions across Healthcare and Enterprise Software.

Contact

Skills

Core Skills

Natural Language Processing (nlp)Machine LearningComputer Vision

Other Skills

AWS EC2AsyncioBM25Beautiful SoupC++Deep LearningDjangoDocker ProductsFastAPIGenaiGenai agentsGenerative AIGraph DatabasesJavaJavaScript

About

passionate about building intelligent systems that shape the future. From data wrangling to model deployment, I thrive on turning complex algorithms into practical solutions.

Experience

1 yr 10 mos

Total Experience

7 mos

Average Tenure

10 mos

Current Experience

Mobikwik

Data Scientist (GenAI)

Aug 2025 – Present · 10 mos · Gurugram, Haryana, India · On-site

Neuriot labs

Data Scientist(GenAI)

Feb 2025 – Aug 2025 · 6 mos · Noida, Uttar Pradesh, India · Hybrid

Oracle-SCM Agent: Developed an enterprise agentic system using LangGraph, OCI, FastAPI enabling sales teams to
query Oracle SCM via natural language.
Built dynamic context filtering, functional-view mapping, routing logic, and HTML report generation for complex enterprise
workflows.
Integrated the multi-agent system with a Java backend enabling seamless orchestration between LLM agents, SCM data sources,
and REST APIs.
VLM–LLM Document Extraction System: Developed a multimodal extraction system for handwritten and poorly scanned
documents using SGLang-hosted VLMs, on-prem LLMs, and a hybrid pipeline combining regex rules, scoring-weight heuristics,
field validators, and VLM predictions to achieve high-accuracy structured data extraction from noisy inputs
Created FastAPI service to consumes requests in batches from Kafka queues and processes them asynchronously using Asyncio.
Enterprise-RAG Framework: Built a modular RAG architecture with agentic orchestration, monitoring, hybrid retrieval, and
interchangeable LLM/embedding pipelines using LangChain, LangGraph, Qdrant.

LangGraphOCIFastAPIJavaregexSGLang+7

Jellyfish technologies

2 roles

Software Engineer (GenAI & CV)

Aug 2024 – Feb 2025 · 6 mos · Noida, Uttar Pradesh, India · On-site

Prior-Authorization System: Built a medical document validation engine using FastAPI, Qdrant, Knowledge Graphs, BM25,
medical embeddings to check compliance with medical policies and generate approval/denial reasoning.
SKU-Auditing System: Developed a YOLO-based computer vision pipeline that automated SKU auditing and reduced manual
effort by 75%.

FastAPIQdrantKnowledge GraphsBM25medical embeddingsYOLO+2

Software Engineer (GenAI & NLP)

Feb 2024 – Aug 2024 · 6 mos · Noida, Uttar Pradesh, India · On-site

Conversational-AI Chatbot: Designed and deployed an LLM-powered chatbot with SQL + Redis integration, improving customer
engagement by 40%.
Implemented Redis-based vector caching and optimised inference pipelines, reducing latency by 30%.

SQLRedisLLMvector cachingNatural Language Processing (NLP)

Utopiic

Data Scientist (NLP)

Aug 2023 – Oct 2023 · 2 mos · Faridabad, Haryana, India · Hybrid

Built and deployed scalable OCR and Deep learning learning models for extracting structured data from electricity
bills and invoices, utilizing computer vision techniques in PyTorch.
Collaborated with the front-end team to integrate backend APIs for real-time sentiment and text analysis, utilizing
transformer-based models for news and legal data.
Enhanced the scalability of machine learning pipelines by deploying them on AWS EC2, enabling real-time
processing of multimodal datasets.

OCRDeep LearningPyTorchtransformer-based modelsAWS EC2Machine Learning