Looking for a job

Akshay Patel

@akshaypatel4

Message

Data Scientist and AI engineer driving production-ready ML solutions.

India

Message

What I'm looking for

I seek roles building production ML/AI systems where I can lead model development, productionization, and create measurable business impact in collaborative, growth-oriented teams.

I am a data scientist and AI engineer with four years of experience delivering production machine learning and agentic solutions that turn complex data into business value.

At IBM I developed an object-detection model for piping and instrumentation diagrams and engineered data pipelines and agent platforms (Agentrix) that showcased advanced agent capabilities and enabled partner lead generation.

At HDFC Life I built serverless document-classification pipelines, a face-recognition KYC application, and an intelligent multi-platform chatbot that handled over two million queries with 97% comprehension, while cutting costs and response times dramatically.

I specialize in deploying scalable, efficient ML systems using tools such as watsonx.ai, PyTorch, AWS Lambda, IBM Cloud, GCP, and container/orchestration platforms, and I focus on delivering measurable business impact through rigorous validation and production engineering.

Experience

Work history, roles, and key accomplishments

Current

AI Engineer

Current

IBM

Dec 2023 - Present (2 years)

Developed and trained an object-detection model to extract 6 component types from Piping and Instrumentation Diagrams, converting unstructured diagrams into structured data and improving downstream analytics; engineered data pipelines and agent platforms that drove a million-dollar opportunity and accelerated partner lead generation.

Object Detection Data Pipelines RAG LLM PyTorch PEFT IBM Cloud DevOps Python

Data Scientist

HDFC Life

Nov 2020 - Dec 2023 (3 years 1 month)

Built serverless document-classification pipelines using AWS Lambda/ECS reducing costs by 85%, developed KYC face-recognition and deployed an intelligent chatbot handling 2M+ queries with 97% comprehension, cutting response turnaround from 24 hours to 10 minutes.