Cascade

Making Autonomous Intelligence Safe & Reliable

Winter 2026ActiveB2BInfrastructureArtificial IntelligenceB2BSecurityInfrastructureTrust & SafetySan Francisco, CA, USA

Company

Cascade builds software infrastructure for autonomous intelligence with self-improving safety and reliability models that continuously maintain, update, and prevent threats and failures at scale.

Verdict

High Signal

Market Opportunity

AI safety and reliability infrastructure is a massive and fast-growing space — every enterprise deploying autonomous agents needs this. B2B infrastructure play with a clear ICP (companies running AI agents at scale). Monetization path via SaaS or usage-based pricing for monitoring/safety tooling is well-established.

Medium Signal

Founder Signal

Both founders are recent UC Berkeley CS grads with BAIR Lab research experience in AI safety and agent systems. Haluk has meaningful industry exposure: 7-month Netflix contract on GenAI Labs, 4-month Amazon internship building agentic AI workflows for Rufus. Adam's industry experience is limited to a NASA data analytics contract and an embedded systems internship, both under 1 year. Strong academic pedigree but neither has shipped a commercial product or held a senior engineering role.

Low Signal

Competition

No competitor data was returned, but the space is heating up fast. LangSmith (LangChain), Arize AI, Weights & Biases, Patronus AI, and Galileo all compete in AI observability and reliability. Anthropic, OpenAI, and Google are building safety tooling natively. Differentiation from research background is plausible but not yet demonstrated in product.

Low Signal

Product

Website shows only a tagline ('Operational Intelligence safe and personalized'), a 'Request Access' button, and a link to a research publication. No product demo, no pricing, no named customers, no API docs, no usage metrics — pure pre-product research-lab aesthetic.

OverallC Tier

Cascade is a research-first team with strong BAIR credentials and a well-timed thesis — AI agent safety infrastructure is genuinely needed at scale. However, both founders are fresh Berkeley grads with no senior engineering or commercial product experience, and the product is pre-launch vaporware with zero visible traction, customers, or revenue. The competitive landscape is crowded with better-resourced players already in market. The core risk is execution: can a research-oriented team build and sell a commercial infrastructure product before better-funded competitors own the space? Worth watching but not yet investable at conviction.

Active Founders

Adam AlSayyad

Founder

Adam is the Co-Founder and CEO of Cascade. Previously, he was a researcher at the Berkeley Artificial Intelligence Research (BAIR) Lab, where his work focused on graph reasoning models, and agentic safety under some of the world's leading ML and AI safety researchers. Adam studied Computer Science at UC Berkeley.

Haluk Cem Demirhan

Founder

Haluk is the Co-Founder and CTO of Cascade. Previously, he built production monitoring infrastructure and scaled agent systems at companies like Netflix and Amazon. His research at BAIR Lab covered long-horizon memory optimization and failure mode taxonomies for AI agents. Haluk studied Computer Science and Mathematics at UC Berkeley.

Cascade

TierC Tier

BatchWinter 2026

Team Size2

StatusActive

LocationSan Francisco, CA, USA