Expanse

Unlock wasted GPU capacity.

Spring 2026ActiveB2BInfrastructureAIOpsDeveloper ToolsEnterprise SoftwareInfrastructureSan Francisco, CA, USA

Company

Expanse unlocks wasted GPU capacity. We recover idle compute through three capabilities: resource prediction (right-sizing job submissions before they reach the scheduler), optimisation suggestions (code and config changes researchers can apply themselves), and failure prediction (catching jobs that will fail before they consume hours of GPU time). We’re four engineers. We ran HPC and GPU training workloads at the largest quant funds and national supercomputing centres. We faced this problem first hand and the only fix was to over-provision and burn millions. Ismaeel built the first multimodal HPC resource predictor as research at EPCC (Edinburgh’s Parallel Computing Centre), which beat every published baseline. This is the tool we wish we had.

Verdict

High Signal

Market Opportunity

GPU compute waste is a massive and growing problem — enterprises, quant funds, and national labs are burning millions on idle or failed compute. B2B infrastructure play targeting HPC centers, AI labs, and financial institutions with GPU clusters. As GPU costs continue rising (H100s at $2-3/hr), even 10-20% efficiency improvement on large clusters represents millions in savings, creating clear ROI-based monetization.

High Signal

Founder Signal

Four engineers with directly relevant hands-on experience: Ismaeel ran large-scale ML at QRT and built a multimodal HPC resource predictor at EPCC that beat published baselines (legitimate research contribution). Nikodem managed GPU cluster platforms at Millennium Management. Yafet and Eren both worked at G-Research on HPC/infra tooling and built novel GNN-based cluster prediction systems. All four have Edinburgh CS degrees plus top-tier quant finance and national supercomputing experience — this is a deeply domain-credentialed team building in their exact wheelhouse.

Medium Signal

Competition

No competitor data returned in search, but the space includes scheduler-adjacent tools (IBM Spectrum LSF, Slurm plugins), cloud cost optimization players (Anyscale, Run:ai, CoreWeave tooling), and observability platforms (Weights & Biases, Grafana-based stacks). Expanse's multimodal prediction approach and failure-prevention angle (vs. just monitoring/reporting) is a meaningful differentiation, and the EPCC research origin provides a proprietary baseline. Big cloud providers could theoretically build this but haven't prioritized HPC-specific prediction.

Medium Signal

Product

Three described capabilities (resource prediction, optimization suggestions, failure prediction) are technically specific and well-articulated, suggesting real engineering work. However, no customer logos, pricing page, live demo, revenue metrics, or named testimonials are present. The website (expanse.sh) and YC description indicate early-stage product with no visible traction signals.

OverallA Tier

Expanse has one of the strongest founder-problem-fit profiles in this batch: four engineers who literally built and ran these exact systems at elite quant funds (QRT, Millennium, G-Research) and national supercomputing centers (EPCC), and at least one has peer-reviewed research beating published baselines in the core problem domain. The market need is real and expensive — GPU waste at scale is a well-documented pain point with direct, measurable ROI. The main weakness is lack of visible traction or customer evidence; this is purely team and domain credibility at this stage, with no revenue signals or named design partners surfaced. If they can convert their insider network at hedge funds and HPC centers into paying pilots quickly, this could be an S-tier — right now it's a strong A waiting on execution proof.

Active Founders

Ismaeel Bashir

Founder

Ismaeel is co-founder and CEO of Expanse. Built the first multimodal HPC resource predictor at EPCC (Edinburgh’s Parallel Computing Centre), beating every published baseline. Previously: ran large scale ML models at one of the world’s largest quantitative funds (QRT). Studied Computer Science at the University of Edinburgh.

Nikodem Bieniek

Founder

Niko is co-founder and CTO of Expanse. Trained and optimised speech recognition models on GPU clusters. Previously: managed the platforms researchers and engineers depended on at one of the world’s largest hedge funds (Millennium). Studied Computer Science at the University of Edinburgh.

Yafet Melake

Founder

Yafet is co-founder and COO of Expanse. Built the first GNN-based cluster graph network for predicting SLURM queue wait times at EPCC (Edinburgh’s Parallel Computing Centre). Previously: tooling and infrastructure for researchers at one of the world’s largest quantitative funds (G-Research). Studied Computer Science at the University of Edinburgh.

Eren Mendi

Founder

Eren is co-founder and CPO of Expanse. Built state-of-the-art decentralised foundation model training systems and performance models. Previously: prototyped emerging technologies in quantitative finance (G-Research). Studied Computer Science at the University of Edinburgh.

Expanse

TierA Tier

BatchSpring 2026

Team Size4

StatusActive

LocationSan Francisco, CA, USA

Disagree with this rating? Let us know

$200 GPU creditsClaim yours →