AC

Ajay Chodankar

AI Researcher & Software Engineer

Agentic AI Systems, Computer Vision, and Cloud Technologies

View My WorkGet In Touch
View / Download Resume
About Me

A research mindset with an engineer's delivery discipline

Ajay Chodankar portrait

AI/ML Engineer & Data Scientist

I am a Data Science Master's Graduate (TU Braunschweig) with robust experience across AI/ML, Cloud Computing, and DevOps, integrating academic research with professional practice.

Specializing in GenAI, I architect Agentic systems and RAG pipelines using Microsoft AI Foundry and Google Gemini. My award-winning AI applications, such as medical data simplifiers and Image-to-BIM pipelines, leverage multi-agent frameworks to achieve high-quality results.

My AI work is backed by Cloud & DevOps expertise. Deploying scalable services on Microsoft Azure, orchestrating with Docker, and utilizing Ansible (IaC), I focus on shipping systems that are reliable after the prototype stage.

With strong intercultural communication skills, I am eager to apply this engineering blend to drive digital transformation in industrial AI, IoT, and smart manufacturing.

GenAI Systems

RAG pipelines, multimodal workflows, and agent orchestration grounded in production realities.

Cloud & DevOps

Azure deployments, Dockerized services, and automation that cut manual setup and friction.

Hackathon Wins

Multiple award-winning prototypes across healthcare and AEC, built under real-world constraints.

Cross-cultural Delivery

Comfortable working across research teams and multilingual environments in Germany and India.

Skills

The stack behind the ideas

The original site's skill matrix worked because it made the portfolio feel multidimensional. I kept that rhythm and made it dynamic-friendly.

AI & Generative AI

Agentic AI SystemsLLMs (GPT-4Gemini)RAG PipelinesVector EmbeddingsMultimodalityVLM (Vision Language Models)Prompt EngineeringChain-of-Thought

Frameworks & Libraries

PyTorchOpenCVPandasNumPyYOLOCrewAILangChainFastAPI

Backend & Engineering

Python (Expert)RESTful API DesignMicroservicesClean Code PrinciplesUnit TestingCARLA

Cloud & DevOps

Microsoft AzureGoogle CloudAmazon Web ServicesKubernetesDockerGit & CI/CDAnsibleSQLLinuxGrafanaTableau

Domain Knowledge

Generative AI (LLMsRAG)Natural Language ProcessingData EngineeringCloud/DevOps Automation | Microservices3D Environment AnalysisComputer VisionAEC/BIM AnalyticsSimulation & Autonomous DrivingAudio Processing (STT/TTS)

Strengths

Analytical ThinkingAdaptabilityEye for DetailAcademic ResearchCoordination
Experience

Research, engineering, and shipping under constraints

04.2024 - Present

Wissenschaftliche Hilfskraft - DevOps & System Administrator

IFAS - Institut fur Flugantriebe und Stromungsmaschinen, TU BraunschweigBraunschweig, Germany
  • Manage lab and project infrastructure end-to-end, from provisioning and build standardization to automated deployments and day-to-day Linux/Windows reliability.
  • Cut infrastructure deployment time by ~70% by standardizing provisioning with Ansible (IaC) across Ubuntu, Rocky Linux, and Windows; reduced manual setup effort by ~80%.
  • Automated recurring operations (DB updates, internal site refreshes, housekeeping) with Python scripts, reducing manual intervention and errors.
  • Hardened delivery workflows with GitLab CI/CD and Docker, adding automated tests and repeatable builds for safer, more frequent releases.

06.2024 - 02.2026

Wissenschaftliche Hilfskraft - AI Researcher

IKE - Institut für Konstruktives Entwerfen, Industrie- und Gesundheitsbau, TU BraunschweigBraunschweig, Germany
  • Build and ship ML features for AEC/VR product. Deliver retrieval-backed assistants, geometry/quality checks, and containerized APIs that integrate with existing CAD/Unity workflows.
  • Improved LLM response quality and consistency by integrating Azure AI Foundry APIs, iterating prompts with test sets, and adding lightweight evaluation/QA checks.
  • Built a CAD-to-Unity VR automation pipeline (Python, FastAPI, Docker) that runs vision-based object recognition and rule-based mapping to assign real-world properties and auto-attach Unity scripts, producing interactive scenes instead of static imports.
  • Detected early-stage design conflicts by analyzing collision/geometry data and generating reports that surfaced spatial issues earlier in the project lifecycle.

11.2023 - 01.2024

Wissenschaftliche Hilfskraft - Robotics

IMN - Institut für mobile Maschinen und Nutzfahrzeuge, TU BraunschweigBraunschweig, Germany
  • Conducted in-depth research on hardware requirements for agricultural mobile robots tailored to specific terrain conditions.
  • Researched about geospatial data acquisition from sensors for effective localization and mapping using QGIS software.

06.2022 - 08.2022

Graduate Apprentice - Data Analyst

LTI - Larsen & Toubro InfotechMumbai, India
  • Hands-on project work in data engineering and cloud. Built batch/stream pipelines and dashboards, and validated results with basic ML and statistical checks.
  • Developed automated ETL pipelines using PySpark, Kafka, and Hadoop to simulate high-volume data streaming and real-time processing.
  • Executed cloud data engineering workflows on GCP and AWS, managing the end-to-end ingestion and cleaning of large-scale structured and unstructured datasets.
  • Built interactive Tableau dashboards and applied ML algorithms to validate statistical models, transforming raw lab data into actionable business insights.

01.2019 - 05.2022

Autonomous Driving Subsystem Member

VCET Solecthon (Autonomous Solar Vehicle R&D)Mumbai, India
  • Pioneered the development of a modular Autonomous Driving System
  • Implemented real-time Computer Vision pipelines using YOLO to detect objects
  • Programmed and optimized low-latency algorithms on Nvidia Jetson and Raspberry Pi
Projects

The same bold project storytelling, now fully data-driven

Click any project for a better case-study view with media, highlights, and clean formatting from the admin panel.

Education

Academic grounding behind the applied work

Achievements

Achievements & Certifications

Best LLM Idea

Healthcare Hackathon 2024 Berlin (UKSH)

Solves a BIG AEC Problem

AEC Hackathon Munich 2024

Published Research

ICOEI 2022 Facial Recognition Security System

References

What collaborators said about the work

Featured
From LinkedIn recommendation

“I met Ajay at a hackathon when he joined my team. The next day, we won the hackathon, in part thanks to his contribution of knowledge about autonomous cars to our hospital logistics challenge. Since then, it has been a pleasure working with him at university and also for TANGILITY, where he has proven himself to be highly competent and responsible. He has a thorough knowledge of various topics and skills that he uses to develop Al applications, multi-agent solutions, and retrieval-augmented generation (RAG). His skills include Python, vision language models, C#, SQL, PostgreSQL, Dockerization, virtualization, large language models (LLM), deep learning, automation, cloud computing, machine learning, Git, and Microsoft Azure. As part of our project, he also taught himself Unity. Ajay works well within a team and doesn't hesitate to research unfamiliar topics.”

Lukas Adrian Jurk

Lukas Adrian Jurk

Architectural researcher, CEO & co-founder @TANGILITY | Experience blueprints with wellplaced in VR/AR before detecting planning issues on the construction site | Avoid operating costs for major construction projects, Tangility GmbH

Dec 19, 2025

From LinkedIn recommendation

“Ajay worked as a HiWi with me for around one year and consistently showed his ability through a variety of tasks. He worked confidently in both Linux and Windows environments and contributed effectively through automation, scripting, and structured use of Git. His combination of research ability and practical implementation—ranging from web scraping and data handling to CAD and meshing—was particularly valuable. He is a reliable and capable team member whom I would recommend for technical and engineering roles.”

Nima Fard Afshar

Nima Fard Afshar

Research Assistant, IFAS, Technische Universität Braunschweig

From LinkedIn recommendation

“Ajay is a skilled Al engineer who builds data ingest pipelines that work reliably, and understands algorithms and LLM/Vision models in depth. I'm impressed by how quickly he picks up new tools and coding languages, diving into the Unity 3D engine with minimal guidance and producing solid code, including for an autono- mous driving project where his knowledge on the state of the art was invaluable. He is easy to collaborate with, especially when bouncing ideas back and forth, and his expertise and reliabil- ity make him a great teammate!”

Marcel Purwins

Marcel Purwins

🏥📐 Wir machen Baupläne in VR und AR erlebbar – damit Fehler nicht erst auf der Baustelle auffallen | Co-Gründer TANGILITY, Tangility GmbH

Dec 19, 2025

Contact

Let's Connect!

I'm always open to discussing new projects, creative ideas, or opportunities to be part of your vision.

[email protected]+49 176 764 66955Braunschweig, Germany