Open to opportunities

GAURAV RAUL

@gauravraul

Message

GenAI engineering lead turning enterprise needs into reliable agentic systems.

India

Message

What I'm looking for

I’m looking for an enterprise AI role where I can build agentic GenAI systems end-to-end—grounded retrieval, rigorous evaluation, and production-ready deployment—with a team that values reliability, security, and cost-aware low-latency inference.

I’m an Applied AI and GenAI engineer with 8+ years of experience translating ambiguous enterprise problems into production software across financial services, telecom, and automation. I stay hands-on from experimentation and model adaptation through APIs, cloud deployment, CI/CD, and observability.

At Fractal Analytics, I architected an Assist framework for reusable enterprise multi-agent applications—defining composable patterns for agents, tools, memory, workflows, and evaluation. I built a FlowAssist natural-language-to-DAG compiler that converts requirements and code into editable visual workflows with node-level orchestration.

I also delivered agentic reconciliation and financial root-cause-analysis capabilities, including “next best action” agents that surface transaction anomalies and propose auditable remediation paths. In parallel, I engineered a hybrid platform that combines distributed batch pipelines with low-latency online inference, fine-tuning Phi-3 and optimizing Gemini for cost-aware, high-accuracy serving.

Earlier, I built production GenAI copilots using RAG and Gemini, and created domain-specific financial intelligence systems powered by transformer fine-tuning. My professional ethos is practical applied research—grounded retrieval, strong evaluation, and engineering discipline—so enterprise AI is reliable, secure, and measurable.

Experience

Work history, roles, and key accomplishments

Current

Senior Data Scientist

Current

Fractal Analytics

Sep 2024 - Present (1 year 10 months)

Architected an enterprise multi-agent Assist Framework and built a FlowAssist natural-language-to-DAG compiler for inspectable, executable workflows. Developed financial wealth management and reconciliation agents, and optimized low-latency model performance using fine-tuned Phi-3 and Gemini 2.0 Flash with context compression.

Multi Agent Systems Text to DAG Workflow Compilation Node Level Orchestration Model Fine Tuning Phi 3 Optimization

Senior Data Scientist

Fractal Analytics via Meraki Talentworks Pvt. Ltd.

Feb 2024 - Sep 2024 (7 months)

Delivered a production retail GenAI copilot using RAG and Gemini 1.5 for real-time conversational summarization, intent classification, grounded response generation, and contextual recommendations. Built end-to-end retrieval and prompt orchestration with focus on latency, context quality, and production reliability.

RAG Gemini Conversation Summarization Reliability

Data Scientist

Cognext.ai

May 2021 - Feb 2024 (2 years 9 months)

Architected financial sentiment-intelligence services using Python, PyTorch, and Flask, including CFSAM, a domain-specific fine-tuned transformer language model. Built natural-language-to-SQL and multi-document intelligence workflows with LangChain and ChromaDB, and operationalized regulated document-intelligence pipelines as reusable APIs and services.

Python PyTorch Flask CFSAM Lang Chain ChromaDB Document Intelligence Detectron2 Image Processing Evaluation Deployment

Manager – Data and Technology

Ipredictt Data Labs Pvt. Ltd.

Jan 2020 - Apr 2021 (1 year 3 months)

Led development of NLP-based lead prediction and recommendation systems, exposing models via Flask APIs deployed on AWS. Built and launched Django-based products (Careerletics Genius and JobCoach) and directed technical strategy with dashboards, data pipelines, and workflow automation.

NLP Recommendation Systems Flask API AWS Lambda AWS RDS AWS IAM Django Python SQL System Design Cloud Infrastructure Third Party Integrations Dashboarding Data Pipelines Workflow Automation

Software Developer

Margo Networks Pvt. Ltd.

Apr 2018 - Jul 2019 (1 year 3 months)

Developed full-stack microservices, REST APIs, and enterprise dashboards using Java, Spring Boot, and Angular 7, collaborating across product and engineering teams. Prototyped a collaborative-filtering recommendation engine as an early foundation for personalization systems.

Java Spring Boot Angular 7 Microservices REST APIs Collaborative Filtering Recommendation Systems