Hello! I’m Gopichand Madala.
An AI engineer focused on agentic systems, model training, and inference. I completed a BE (Hons) in Computer Science (AI & ML) at Chandigarh University with a CGPA of 8.04.
Research at Samsung R&D covered diffusion denoising for endoscopy images and policy checks for the Samsung Finance+ chatbot. As an IndiaAI Fellow, I fine-tuned DeBERTa to estimate product volume from unstructured catalog text.
At ZysecAI I built document ingestion for RAG and replaced a 70B model with a fine-tuned 1.5B coder model. At PurpleMerit I built a gateway that routes prompts across model tiers and cut average inference cost by 27%.
I speak English, Telugu, and Hindi.
Experience
Education
- Chandigarh UniversityBE (Hons) Computer Science (AI & ML), 8.04 CGPA•Aug 2021 – May 2025
Achievements
- Academic Excellence AwardChandigarh University•April 2025
- IndiaAI Fellow1 of 150 selected across India•Jan 2025
- Amazon ML ChallengeRanked 12th of 26,000•April 2023
- CISCO & NASSCOM HackathonWinner of a 4-day hackathon•Feb 2023
- Chandigarh Police HackathonWinner•Nov 2022
- Merit Scholarship50% scholarship through undergrad•Sep 2021
- DMLS Book ClubTeaching assistant with Chip Huyen•April 2024 – June 2024
What I do
Agentic systemsLLM inferenceModel trainingRAGPyTorchFastAPI
Stack
Python
PyTorch
Hugging Face
FastAPI
Docker
Kubernetes
PostgreSQL
Redis
Linux
GitHub
Let’s connect
Open to research, inference work, and teams building agentic systems. Reach out.