DevTools · AI & ML · 200–1000 · Asia-Pacific #b038_02

Context

We design our own accelerator chips, run private supercomputing clusters, and build domestic language models and industrial applications for manufacturing, materials and healthcare partners.

Industry

DevTools & Software · AI & ML

Size

200–1000 employees

Region

Asia-Pacific

Tools budget

$5K–20K per month

Needs

Cloud security posture. Our private clusters and cloud tenants are configured by different teams, and we have no single view of misconfigurations.

Today: separate teams configure in-house clusters, with certified security management only for the products division · Urgency: later

Logs & metrics at controlled cost. Telemetry from thousands of accelerators overwhelms our monitoring stack, and retaining it all has become expensive.

Today: in-house telemetry around self-built supercomputers and proprietary accelerators, tuned by hardware teams · Urgency: now

LLM evaluation & testing. We need repeatable evaluation of our local-language models against domain benchmarks before each release to partners.

Today: internal benchmarking of home-grown language models, extended through multimodal consortium work · Urgency: now

Who decides

vp engineering · security review

Expects from suppliers

Works across on-premise hardware and public cloud; Handles very high cardinality hardware telemetry; Clear data residency guarantees

Not a fit

Pure SaaS tools that cannot run on our own hardware; Evaluation suites limited to English benchmarks

Stack

on-prem / private cloud, Kubernetes, multi-cloud

Buying triggers

scaling transaction volume, new market or product launch