> Markdown version of [/@purnima-kumari](https://www.wearedevelopers.com/@purnima-kumari). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Purnima Kumari
Senior Solutions Architect at DigitalOcean optimizing AI and cloud architectures.

## About

Purnima Kumari is a Senior Solutions Architect at DigitalOcean. She focuses on translating complex cloud and AI architectures into clear, practical systems for engineering teams.

Rather than just throwing more hardware at a problem, Purnima looks at the underlying system design. She frequently speaks on scaling LLM inference by rethinking parallelism, memory layouts, and runtime configurations. Her work dives into vLLM internals and KV cache optimization to double production throughput without needing extra GPUs.


## Upcoming Sessions

### Conference India 2026 · November 25, 2026
Bengaluru, India
- [How 8 GPUs Became 16: Rethinking LLM Inference Through Parallelism, Memory, and Runtime Design](https://www.wearedevelopers.com/events/conference-india-2026/sessions/2003-how-8-gpus-became-16)





## Links

- [LinkedIn](https://www.linkedin.com/in/purnima-kumari18/)