Top AI Product · Jun 2026
Coverage described Salus as a YC W2026 startup for production agent workflows where mistakes are costly, and said it sits between agents and tools to block unsafe actions and return actionable feedback.
Guardrails to validate your agent's actions before they execute
Public company, workplace, funding, and market signals
Updated Jul 29, 2026
Salus is a YC W26, 2026-founded AI-agent infrastructure startup based in San Francisco/Palo Alto that puts a policy-aware runtime between agents and tools so proposed actions are validated against evidence and policy before execution. Public company/LinkedIn pages report 2 employees and $625k raised across two rounds; no public CEO is named, while Kevin Pan and Vedant Singh are the visible founders.
Primary product
Policy-aware runtime guardrails / enforcement API for AI agent actions before they execute.
Founded
2026
Headquarters
San Francisco, California, United States
Team size
1-10 employees
Industry
Software Development
Sub-industry
AI agent infrastructure / runtime guardrails
Offices
0 jobs at Salus
Check back later for new openings
Business model
Investors
Public pages describe a tiny, early team with a high bar for rigor, precision, and simple mechanisms. They emphasize direct founder communication, no cover letters or portals, shadow testing, and shipping on the hot path; titles are described as something to be invented later.
Compensation
Public hiring copy only exposes one posted comp signal: a Founding GTM / BDR role at $35-$45/hour. Broader compensation and benefits are not publicly disclosed.
Pricing
Not publicly posted; demo-led/contact-sales model with subscription software and runtime usage/metering features.
Differentiators
Technology
Customers
Competitors
Top AI Product · Jun 2026
Coverage described Salus as a YC W2026 startup for production agent workflows where mistakes are costly, and said it sits between agents and tools to block unsafe actions and return actionable feedback.
StartupHub.ai · May 2026
An external teardown/coverage piece described Salus as a policy-aware runtime proxy for agents, focused on action validation, evidence caching, structured retries, and safety features such as PII, budget, and loop protection.
Y Combinator on LinkedIn · Feb 2026
YC highlighted Salus as runtime guardrails for AI agents, emphasizing proactive action validation, evidence-backed decisions, policy checks, and self-correcting feedback at runtime.
Y Combinator
Official launch page describing Salus's runtime validation, evidence cache, policy checks, and benchmark claims, including self-repair on blocked actions.