Lilidi AI Awarded $10,000 in Crusoe Cloud Credits to Power Lilidi Twin

Lilidi AI has been awarded $10,000 in cloud credits from Crusoe, the AI infrastructure company behind Crusoe Cloud. Crusoe's platform has two sides: Infrastructure Cloud, its GPU compute, storage and networking, and Intelligence Foundry, its service for running and customising open AI models.

For a company building AI employees, compute is not a side cost. Every conversation a Lilidi Twin holds is followed by work nobody sees: a summary of what was said, a lead record, a draft follow-up for the owner to approve, and new facts filed into the Twin's knowledge. The credits let us move more of that work onto infrastructure designed for AI from the ground up.

Who Crusoe is

Crusoe builds what it calls energy-first AI infrastructure, with NVIDIA and AMD GPUs for training and inference at scale. Crusoe Cloud offers GPU instances, managed Kubernetes and Slurm clusters, and S3-compatible storage. Intelligence Foundry adds serverless inference for open-weight models through an OpenAI-compatible API, serverless fine-tuning, and dedicated model deployments on H100 and H200 GPUs.

What the credits will power

The work after every conversation

Summaries, lead notes, follow-up drafts and knowledge extraction run on large language models today. Through Intelligence Foundry we will test open models such as Llama and DeepSeek for these jobs, measured on quality, speed and cost against what we use now. Anything that replaces a current model has to write at least as well, in every language our customers speak.

Faster answers from the Twin's knowledge

A Twin answers from the website, documents and FAQs its owner approved. We will use Crusoe GPUs to test how we index and search that knowledge, so the right fact reaches the conversation sooner.

Careful experiments with fine-tuning

We will test fine-tuning open models on Lilidi's own materials and on conversations our team writes for that purpose. We do not train shared AI models on our customers' conversations, documents or recordings, and the credits do not change that.

What stays the same

The live voice and the video face of a Twin keep running on the providers our customers use today. Calls, minutes and plans do not change.

Why it matters for Lilidi Twin customers

AI employees are judged in seconds: how fast they answer, how well they remember, how good the follow-up is. Better infrastructure behind the conversation shows up as a shorter wait after a call, sharper summaries and lower running costs, which keeps plans starting at $19 a month realistic as usage grows.

A pattern, not a one-off

The Crusoe credits join the support Lilidi AI already has from the Google for Startups Cloud Program and Microsoft for Startups, an invitation to the Web Summit ALPHA programme and selection for the TechBBQ Startup Showcase. Next week the team is at TechCrunch Disrupt 2026 in San Francisco, 13-15 October, where you can talk to a live Twin and see what an AI employee does after the call.

FAQ

Try it

Talk to a live Twin, or build your own from your website and documents.

Crusoe, Crusoe Cloud and Intelligence Foundry are trademarks of their respective owner. Cloud credits do not imply investment, partnership or endorsement.