graph by datavruti
mygraph.id/adarshdipakbhai37

Adarsh Dipakbhai Pala

AI Engineer

Bengaluru, IndiaBuilding since 2022 (~5 yrs)
Headline

My path runs from academic research in medical imaging and predictive modeling to building agentic AI tools now adopted into real engineering workflows.

60engineers using the assistant

engineers

97.2%P&ID extraction model accuracy

mAP@50

107automated tests in CI/CD pipeline

tests

3ReAct agents coordinated in CAD assistant

agents

Represented Roles

Specialized Functions & Roles.

Engineering capabilities and specialized roles demonstrated across real production work.

AI Engineer

Primary Role

Applied AI Researcher

Computer Vision Engineer

MLOps Engineer

Signals

Live in Production.

Each signal ties back to a specific, evidenced moment.

Built

4 moments

Systems shipped end-to-end, from schema to deploy.

  • •Built an agentic CAD assistant coordinating three ReAct agents
  • •Shipped the agent service with a five-stage CI/CD pipeline
  • •Trained YOLOv11 to hit 97.2% mAP on P&ID extraction

Solved

2 moments

Gaps closed - false positives, edge cases, dead ends.

  • •Achieved 0.88 Dice score on fetal skull segmentation
  • •Co-authored a tool-life prediction paper with 97% accuracy

Learned

2 moments

New tools and depth picked up on the job.

  • •Exploring evaluation-driven, production-grade AI architectures
  • •Experimenting with LLM agents for tool use and retrieval
Growth trajectory

Adarsh's career, mapped.

A timeline of milestones captured directly from real work.

  1. Published early manufacturing ML research

  2. Pursued independent research in imaging

  3. Moved into technical sales engineering

  4. Added formal ML and PM credentials

  5. Returned to graduate study

  6. Shifted into applied AI tooling

  7. Recognized for internship impact

Stack

Daily tools.

What Adarsh actually reaches for, pulled from the tags attached to captured moments.

Work

Key Work By Role

What they built, shipped and owned at each role, with impact numbers where there are any.

Atlas Copco India Ltd.

AI Intern · 2026 - 2026
May 2026 - Jul 2026

P&ID Extraction System

End-to-end P&ID extraction system training YOLOv11 on 4,305 annotated instruments across 62 drawings, integrating PaddleOCR for text-to-symbol association and hosting the containerized model on Azure Container Apps.

97.2% mAP@5097.6% recall96.3% precisionTrained on 4,305 annotated instruments across 62 drawings
Stack:YOLOv11PaddleOCRComputer VisionAzure Container Apps
View case study
May 2026 - Jul 2026

Agentic AI Assistant for CAD Design

Agentic AI assistant embedded in Autodesk Inventor using LangGraph and Azure OpenAI, coordinating three ReAct agents for fastener selection, sheet-metal design-for-manufacturability, and ASME verification, grounded in three Chroma knowledge stores with human approval before live CAD changes.

107 automated tests22 labelled LLM evaluation cases5-stage CI/CD pipelineCoordinated 3 ReAct agents across fastener, DFM, and ASME workflowsGrounded in 3 Chroma knowledge storesAdopted by 60 engineers as part of their design workflow.
Stack:FastAPIDockerAzure DevOpsCI/CDLLM EvaluationLangGraphAzure OpenAIReActChromaAgentic AI
View case study

Xylem Water Solutions India Pvt. Ltd.

Senior Engineer · 2022 - 2025

Prepared technical proposals for pump tenders

Supported distributors, key accounts, and end customers, partnering with sales managers on pre-bid and post-bid tenders, selecting pumps in line with customer specifications and preparing technical proposals including clarifications and deviation lists.

Stack:Technical ProposalsTenderingPump Selection
View case study
Outside Work

Side projects, open source & publications.

Independent builds, open-source contributions, technical writing, and experiments beyond the day job.

Side project

Fetal Skull Landmark Detection & Segmentation

Independent deep learning research on medical imaging detecting fetal skull landmarks and segmenting the cranium from noisy ultrasound images using custom CNN regression plus U-Net segmentation with ellipse fitting.

Experiment

LLM Agent Tool Use

Exploring how LLM agents can reliably use external tools, retrieve domain knowledge, interpret engineering documents and visual data, and execute constrained actions rather than just generate text.

Experiment

Evaluation-Driven AI Development

Studying evaluation-driven AI development, multimodal models, and production-grade AI architectures to make systems more accurate, traceable, and dependable in real-world applications.

Learning

Formal record.

  • Summer Analytics 2026 (6-week course)
    IIT Guwahati
    2026
  • Gold Alpha Challenge
    WorldQuant
    2026
  • Master of Management, Business Analytics & Technology Management
    Indian Institute of Science (IISc)
    2025 - 2027
  • AWS Certified Machine Learning Engineer – Associate
    Amazon Web Services
    2025
  • Project Management Professional (PMP)
    Project Management Institute
    2025
  • B.Tech, Mechanical Engineering
    Nirma University
    2017 - 2021
Activity heatmap

A record of momentum.

A combined view of Adarsh's activity across the tools used daily.

LessMore

Profile last updated 3 days ago.

Liked Adarsh's graph?

Build yours today.

Start with your resume