Open to opportunities

Gowtham S

@gowthams1

Message

Machine learning engineer specializing in LLM pretraining, data pipelines, and scalable model evaluation.

India

Message

What I'm looking for

I seek ML engineering roles focused on LLM pretraining, scalable data and inference pipelines, and production RAG systems in collaborative, research-driven teams.

I am a machine learning research engineer focused on large language model pretraining, evaluation, and high-performance data pipelines. I have contributed to training and analyzing model-scaling behavior across hundreds of millions to billions of parameters, enhanced architectures with multi-headed latent attention and key-value caching, and led dataset preparation using CC-Net, GoClassy, Ungoliant and custom filters.

I build production-ready tooling and pipelines in Rust, Go and Python — including exact line-level deduplication, MinHash near-duplicate detection, asynchronous Hugging Face-compatible pretraining pipelines, and domain-adapted RAG systems. I have experience finetuning large models (e.g., Qwen-8B), pretraining code models, and deploying retrieval workflows for PDF and conversational RAG applications.

Experience

Work history, roles, and key accomplishments

Current

Machine Learning Research Engineer

Current

BluBridge

Mar 2025 - Present (1 year)

Contributed to training and evaluation of large LLMs (124M–1B+ parameters), conducted comparative experiments between LLAMA 3.1 and OLMA, and applied scaling-law methods to predict model performance while improving architectures and caching to optimize runtime.

PyTorch Jax Hugging Face Scaling Rust Data Pipelines Model Evaluation

Machine Learning Engineer

Zeb

Jun 2024 - Jan 2025 (7 months)

Designed Rust-based data pipelines and deduplication systems and built RAG applications for PDF search and code-generation tools, improving retrieval and developer workflow automation.

Rust fastAPI AWS Vector Databases And Deduplication