|
AI Developer with 3+ years building agentic AI systems and enterprise automation. Previously at SAP Labs, now bringing that experience to grad school and industry opportunities.
The story so far
- years shipping AI and enterprise systems
- 3+
- critical SAP incidents resolved
- 600+
- less support effort with multi-agent workflows
- 30%
- IEEE-published research papers
- 2
Who I am
About Me
~ whoami
Varun Sadashive Gowda
~ cat stack.txt
Python · FastAPI · Next.js
LangGraph · CrewAI · OpenAI
~ cat status.txt
● Open to Summer 2027 Internships
~ ls interests/
→ LLM infrastructure
→ multi-agent orchestration
→ applied ML systems
~ █
Driven AI engineer who transforms complex enterprise problems into scalable intelligent systems. I combine architectural thinking with hands-on development to deliver agentic AI solutions that improve operational efficiency.
Known for rapid learning, clear technical communication, and collaborating across teams to build reliable, production-ready AI systems.
Interested in LLM infrastructure, multi-agent orchestration, applied ML systems.
About
Skills & Expertise
Python
ML pipelines, automation, AI development
ABAP
SAP development & integrations
JavaScript
Frontend interactions & scripting
C / C++
Systems & algorithmic foundations
SQL
Database design & semantic search
Java
OOP & backend development
Also work with
AI & ML
Career
Work Experience
Agentic AI Developer
SAP Labs · Bengaluru · GCID CSM Operations Center & AI CoE EMEA
- Designed and deployed multi-agent workflows for SAP GCID Support engineers, reducing work effort by 30% using CrewAI & LangGraph.
- Developed custom agent tools for data extraction, solution analysis, document parsing (PDF/CSV), automated reporting, and image extraction from PDFs (SAP BOCR).
- Built and maintained an MCP (Model Context Protocol) server integrating CLOKS (data warehouse) and AIM (AI agents) repositories.
- Built a RAG pipeline on Databricks — ingests enterprise documents, converts to Markdown, performs semantic chunking, and indexes embeddings in SAP HANA DB.
- Automated Databricks job execution via GitHub Actions, enabling CI/CD-driven orchestration of data workflows.
- Led Docker containerization efforts, cutting image size by 67% (6 GB → 2 GB).
Associate Solution Support Engineer
SAP Labs · Bengaluru · TS PRC ERP Supply Chain APJ
- Resolved 600+ critical issues for SAP customers through incident management. [ABAP, C++]
- Mentored junior engineers and external hires, enhancing team productivity by 20% through code reviews and technical workshops on Transportation Management.
- Collaborated with EWM, MM, SD, and BN4L cross-functional teams to create datasets for internal systems.
- Compiled a knowledge repository accessed by 300+ SAP customers.
- Assisted customers with greenfield/brownfield implementation design and clean core operations.
AI Intern – Deep Learning & Computer Vision
Sandlogic Technologies · Bengaluru
- Implemented ResNet and YOLOv5 models for supply chain computer vision applications. [Python]
- Annotated datasets and developed a client project end-to-end using YOLOv5.
- Optimized training pipelines through hyperparameter tuning and data preprocessing.
Background
Education
Master of Computer Science (MCS)
CurrentTexas A&M University · College Station, TX
Seeking Summer 2027 internships in AI / ML / Software Engineering
B.E. – Electronics and Communication Engineering
JSS Science and Technology University, India
Computer Networks · Operating Systems · Cryptography & Network Security
Publications
Research Papers
Text Independent Speaker Recognition and Classification using KNN Algorithm
Achieved 96.97% accuracy using a double distance measurement method, surpassing KNN's 84.85%. Investigated speaker idolization for improved transcription and secure smart environments.
Read paperReal Time Theft Detection Using YOLOv5 Object Detection Model
Automated theft detection using YOLOv5, enabling real-time identification of theft events with alerts to owners and authorities. Leveraged deep learning and object detection for real-world deployment.
Read paperWork
Projects
Personal Portfolio Website
This site — a full-stack portfolio with a FastAPI + pgvector RAG backend, streaming AI chat widget, Next.js 16 frontend, and edge-rendered OG images. Deployed on Vercel + Render.
Media Player Control by Hand Gestures
Real-time media player controlled via hand gestures (Play, Pause, Skip, Volume) using OpenCV integrated with VLC.
Perovskite as a Triboelectric Nanogenerator
Fabricated a 5×5×0.1cm thin film and characterized its triboelectric properties to investigate nanogenerator performance.
Say hello
Get In Touch
TAMU MCS student in College Station, open to Summer 2027 internships in AI, ML, and software engineering. Drop a message. I'd love to connect.