🎓 M.S. in Artificial Intelligence @ UC Santa Cruz 🧠 Machine Learning Engineer | NLP Researcher
I'm passionate about building intelligent systems that can reason, learn, and interact naturally. Currently pursuing my M.S. in AI at UC Santa Cruz, focusing on agentic systems, LLM fine-tuning, and multimodal AI.
Right now, I'm building browser agents that understand and automate complex web workflows. I've also worked on training reasoning models with RL (GRPO, DPO), agentic RAG systems for multi-hop QA, and fine-tuning vision-language models for figurative language understanding.
I'm seeking full-time roles starting April 2026 in ML, NLP, or AI research/engineering.
- Large Language Models: SFT, DPO, GRPO, preference tuning, alignment, reasoning traces
- Efficient Training: PEFT, QLoRA, distributed training, FSDP
- Multimodal AI: Vision-language models, VLMs, Image-based QA, CLIP
- Agentic Systems: Browser automation, tool-use, RAG pipelines, reasoning-aware agents
- NLP & IR: Transformers, retrieval-augmented generation, multi-hop QA
Languages: Python, C++, JavaScript, TypeScript, SQL, Bash/Shell
ML/DL Frameworks: PyTorch, TensorFlow, PyTorch Lightning, Hugging Face Transformers, TRL, Ray
LLMs & GenAI: Fine-tuning, VLMs, SFT, PEFT, DPO, QLoRA, RAG, Prompt Engineering, Agent Design
Tools & Libraries: NumPy, Pandas, Scikit-learn, Matplotlib, spaCy, NLTK, Unsloth, Playwright, OpenCV
MLOps & Infra: Docker, Kubernetes, AWS (Lambda, SageMaker, EC2, S3, RDS), GCP, Git, MLflow, W&B
Databases: PostgreSQL, MySQL, MongoDB, Redis, FAISS, Supabase
-
AlterNet-K: A Small and Compact Model for Glaucoma Detection Biomedical Engineering Letters (BMEL), 2023
-
Localization of Lung Abnormalities using Self-Supervised Equivariant Attention AOCR + KSR 2022, Biomedical Engineering Letters (BMEL)
- LinkedIn: linkedin.com/in/gavin-p-dsouza
- Website: gavinds.com
- X: @_gavinds