Two ex-Microsoft engineers built AI teammates that read the policy, check your eligibility, and do the thing - not just point at a PDF.
The most expensive line in an HR budget isn't payroll - it's the interruption. Every "quick question" about PTO, a bonus, a leave of absence, or which health plan to pick pulls a human being off the work that actually needs a human. Cascade AI's wager is that most of those questions never needed a person in the first place. They needed a system that could read the policy and finish the job.
That distinction - reading versus doing - is the whole thesis. The market is crowded with tools that surface a document and wait for someone to act on it. Cascade calls that table stakes. Its agents, the company says, verify who you are, check what you're eligible for, interpret the relevant policy, and then execute the workflow across the systems of record where HR data actually lives: Workday, ADP, UKG, ServiceNow.
The founders came to this problem from an unlikely direction. CEO Ana-Maria Constantin studied astrophysics and computer science before spending years as a software engineer at Microsoft, building large-scale cloud systems. Co-founder Pulak Goyal spent nearly six years in Microsoft's AI and cloud groups. They pointed that systems-thinking pedigree at the least glamorous corner of the enterprise - the HR help desk - which turns out to be exactly the kind of unsexy, high-volume problem that makes durable software companies.
Founded in 2022, the company is small: roughly 22 people, based in Bellevue, Washington, in the Seattle orbit. But it has attracted outsized attention. In October 2024 it announced a $3.75 million seed round led by Gradient, Google's AI-focused venture fund, with participation from Myriad Venture Partners and Success Venture Partners - bringing cumulative funding to about $5.45 million.
Cascade sells AI "teammates," each purpose-built for one operational domain, integrated with the systems of record, and governed by the customer's own policies. Three do most of the work.
Handles the full lifecycle of an HR request - from the first question about benefits, leave, policy or compensation all the way to the closed transaction. HR teams keep the work that needs them and lose the ticket queue that doesn't.
Walks employees through benefits choices with personalized, policy-aware support. It's aimed squarely at open enrollment - one of the largest financial decisions people make each year, usually made alone and guessing.
Resolves IT service requests end-to-end - from access provisioning to incident triage - without a ticket queue standing between the employee and the fix. The same agentic pattern, pointed at the IT desk.
An agentic AI layer that sits across HRIS, ITSM, payroll, benefits and identity systems. Every interaction is verified, compliant, auditable, and scoped to the employee's role and permissions.
The agent identifies the employee and understands who is asking.
It checks eligibility and permissions - what this person can actually access.
It interprets the organization's own policy for the situation at hand.
It runs the workflow in real time across the systems of record - and closes it out.
Cascade's public pitch rests on ticket deflection. The company says customers see an immediate cut in HR inquiry volume that grows as employees adopt the tool - and that some organizations with complex, global teams have made Cascade the primary path for all HR questions.
Cascade is built to run on infrastructure the customer already owns, with out-of-the-box connections to the systems HR and IT already use. HRIS data flows in via Finch. No rip-and-replace, the company says, and no custom engineering to get started.
Note: Cascade AI did not maintain a public YouTube channel at time of writing. Product demos are available by request via the site above.
Cascade AI is a Bellevue/Seattle-based startup building an agentic AI platform for HR and IT operations. Its AI teammates sit across systems of record like Workday, ADP, UKG and ServiceNow to answer employee questions and resolve requests across benefits, leave, compensation, policy and IT - executing workflows in real time rather than just surfacing information. Founded in 2022 by former Microsoft engineers Ana-Maria Constantin and Pulak Goyal, the company raised a $3.75M seed round led by Gradient, Google's AI-focused venture fund, and says customers see roughly a 50% drop in HR ticket volume after deployment.
Last updated: