
1. Technology overview
Expanse is a London company (Y Combinator Spring 2026 batch) that builds software to predict how much GPU, memory and time a compute workload needs before it runs. It targets teams running AI training, HPC simulation and inference fleets.
One command installs Expanse on an existing SLURM or Kubernetes cluster, where it collects telemetry alongside workloads. A model pretrained on HPC and AI workloads adapts to the customer's environment and returns predicted GPU allocation, memory use, runtime and completion probability as a confidence distribution.
2. Highlighted features & specifications
- Pre-run analysis: predicts GPU allocation, memory, runtime and completion probability.
- Cluster install: one command on SLURM or Kubernetes.
- Telemetry: collected alongside existing workloads.
- Learning: pretrained model adapts to each infrastructure.
- Confidence distributions: each prediction includes its uncertainty.
- Delivery: ask in the terminal or apply predictions automatically.
3. Potential application
- AI labs estimating resources before large training runs.
- HPC centres reducing failed simulation jobs.
- Data centre operators improving GPU utilisation.
4. Benefits
Reduces failed and over-provisioned jobs by replacing guesses with evidence. Works on existing clusters without changing workflows. Predictions improve as more workloads run.
5. Market trends & opportunities
GPU capacity is scarce and expensive, and failed training runs waste it. Tools that raise utilisation of existing clusters are in demand as AI infrastructure spending grows, including among Southeast Asian data centre operators.
6. Other information
Y Combinator Spring 2026 company. Launch YC post dated 2026-05-04.
One platform connecting startups, corporations, research institutions, regions and investors











Playbook & Handbook · February 6, 2026
Winning Accelerator Playbook 2026 by BambuUP
RETAIL E-COMMERCE: AI AND REAL-TIME DATA ARE TRANSFORMING HOW RETAIL BUSINESSES OPERATE








































