If vLLM already solved LLM serving, why did SGLang appear?
Meet Neysa at MindTech 2026
Neysa is the Founding Sponsor of MindTech 2026, the two-day summit on mental health in the age of AI, hosted by TANUH at the Indian Institute of Science, Bengaluru. Find us at Booth A2 in the showcase on both days.
MindTech brings clinicians, researchers, technologists, startups and policymakers into one room. We are there for the part of the conversation that often comes last: what a mental-health AI tool runs on, and how that decides whether it is available, responsive and affordable when someone needs it, with patient data kept in India.
Speak with the Neysa team to see how the right infrastructure decisions can help your mental-health AI stay responsive and affordable for everyone who uses it.
Visit Us at Booth A2
Booth No. A2 | Showcase | October 5–6, 2026
See Neysa Velocis, our AI Acceleration Cloud, live: pick an open-weight model from the Model Library and deploy it to a dedicated inference endpoint.
Open pre-configured Jupyter or VS Code workspaces in minutes. Provision the runtime your workload needs: bare metal, GPU VMs, managed Kubernetes (VKE) or dedicated inference endpoints.
Talk to our engineers about serving models with steady response times, and scaling as more people use your tool.
Neysa on Stage
Sujit Janardanan, CMO at Neysa, shares notes from the infrastructure side of building AI for mental health.
Building Health AI in India
Notes from the infrastructure side.
5 October 2026 | 2:00 PM | Main Hall | 20 minutes
Speaker: Sujit Janardanan
Selecting a model, fine-tuning it, and shaping how it speaks to patients is only part of building AI for healthcare. The rest is what the model runs on.
A health tool has to be there whenever someone reaches for it, reply without a long pause, and hold up as more people use it. It also has to do that at a cost that lasts, with patient data kept where it belongs. The session covers how teams building in this space can plan for all of it early, so the infrastructure never becomes the reason the work stalls.
- Cost and reliability. Cost, reliability and where patient data lives matter as much as which model you pick.
- Availability. Someone who needs help will not wait. The tool has to be up when they reach for it, and reply without a pause that feels like being ignored.
- Scalability. The tool has to keep working as more people use it, and the team behind it should be able to move to a newer model when a better one comes along.
- Data privacy. Where patient data has to stay in India, fine-tuning, serving and scaling have to happen in India too.
If you’re building, funding or deploying AI for health, this is a conversation you won’t want to miss.
Join Sujit at MindTech to plan the infrastructure before it becomes the bottleneck.
Coming to MindTech? Let’s Meet.
Whether you’re a startup building a mental-health app, a research team fine-tuning models on clinical data, or a hospital or public-health program deciding where your AI should run, we’d like to understand what you’re working on.
Visit us at Booth A2 or book a meeting with Neysa to discuss your AI workload, availability and response-time needs, where your patient data has to live, and the path from pilot to production.
FAQs
What is Neysa’s role at MindTech 2026?
Neysa is the Founding Sponsor of MindTech 2026, engaging with clinicians, researchers, startups and public-health teams around the infrastructure needed to build and run mental-health AI in India.
Where can I find Neysa at MindTech?
At Booth No. A2 in the showcase, on both days of the summit, October 5 and 6, at the Indian Institute of Science, Bengaluru. On October 5 in the Main Hall, Sujit Janardanan, CMO at Neysa, speaks at 2:00 PM, and Ritesh Dhoot’s session runs between 3:00 PM and 5:00 PM.
What will Ritesh Dhoot’s session cover?
Ritesh will look at building mental-health AI from the infrastructure side: cost and reliability, availability when someone reaches for help, scaling as usage grows and moving to newer models, and keeping fine-tuning, serving and scaling in India when patient data has to stay in the country.
What will Sujit Janardanan’s session cover?
In “Building Health AI in India”, Sujit Janardanan, CMO at Neysa, shares notes from the infrastructure side of building AI for healthcare: planning early for cost, reliability, availability, scale and where patient data lives, so infrastructure never becomes the reason the work stalls.
Why does infrastructure matter so much for mental-health AI?
Someone in distress will not wait. Whether the tool is up when they reach for it, how quickly it replies, what it costs to keep running and where patient data lives are all decided by the infrastructure, not only by the model.
Can we fine-tune and serve our models in India?
Yes. Velocis runs in India, so you can fine-tune, serve and scale your models in the country. You bring the model; we give you the GPUs to train it and the endpoint to serve it.
Can we switch to a newer model later?
Yes. You choose the model, whether open-weight or your own fine-tuned version. When a better one comes along, you can deploy it on the same infrastructure without rebuilding your stack.