Enterprise AI Platform. Simple. Scalable. Open.

Felafax’s AI platform runs on any accelerator—from Google TPU to AWS Trainium, NVIDIA to AMD—achieving 2X cost-efficiency without sacrificing performance.

Request Access Schedule Demo

Key Features of Felafax

Scale Effortlessly
One-click spin-up of clusters from 8 to 1024 TPU chips. Our framework seamlessly handles training orchestration at any scale.
Performance at Lower Cost
Our custom training platform, built from the ground up, uses XLA compiler and JAX. Get H100-level performance at 30% lower cost.
On-prem deployment
We deploy in your VPC, ensuring your data never leaves your network and remains secure and private.
Highly Customizable
Use our no-code UI for fine-tuning or drop into a Jupyter notebook to tailor your training run. Full control with zero compromises.
We handle all ML Ops
We provide optimized model partitioning for larger models like Llama 3.1 405B, handle multi-controller training and inference. Focus on innovation, not infrastructure.
Out-of-the-Box Templates
Choose between PyTorch XLA or JAX. Get started quickly with pre-configured environments and all necessary dependencies installed.

Want to fine-tune and deploy Llama3 in your enterprise VPC?

Please reach out to us, and we'll work with you to get you set up. 🙂

Reach out

Meet our team

Nikhil Sonti
Co-Founder & CEO
Over 6 years at Meta and 3+ years at Microsoft, Nikhil has worked on ML inference infrastructure for Facebook Feed, focusing on performance and efficiency.
Nithin Sonti
Co-Founder & CTO
Nithin has over 5 years of experience at Google and Nvidia, specializing in building large-scale ML training infrastructure. He worked on building the trainer platform for YouTube recommender models and fine-tuned Gemini for YouTube.