July 23, 2026
Observe.AI Signs Multi-Year AWS Deal to Move CX Agents into Production
Observe.AI has announced a multi-year deal with Amazon Web Services (AWS), giving enterprises a route to run its Agentic CX Platform on AWS infrastructure across customer service operations. The agreement covers deployment, and AWS Marketplace will also carry Observe.AI products, so buyers can draw on existing AWS commitments instead of opening a separate procurement process.
The two companies are targeting production environments rather than pilots, and the agreement emphasises reliability, security, compliance and scale. These same four requirements stall AI deployments in regulated service operations long after the technology has proven itself in a demo.
Observe.AI’s own argument is that most contact centres are still running on legacy IVRs, disconnected automation tools, manual agent workflows and analytics that only report on what already happened. Those systems were never designed for software that can understand intent, take action and complete a task on its own.
Three Sets of Agents Doing Three Different Jobs
The platform groups its AI agents into Customers, Frontline Teams and Operations.
Automation Agents handle voice and chat interactions end to end, grounded in enterprise knowledge and connected to systems of record so they can complete the work rather than route it. Companion Agents support human agents before, during and after a conversation by surfacing context, guiding required steps, recommending next actions and taking on after-call work. Operations Agents give team leaders visibility into quality, compliance, intent trends, root causes, agent behaviour and how the automation itself is performing.
The third category is the interesting one. Governance of AI agents is becoming its own workload, and vendors that can evaluate machine and human performance in the same system have an advantage over those still selling analytics as a separate product.
Observe.AI’s coaching layer has been one of the stronger real-time agent assist offers on the market, with users reporting handle time down by around 23%, conversions up roughly 10% and compliance close to 97%.Its quality assurance tools score every interaction rather than a sample of ten calls per agent per month, which is where most manual QA programmes run out of road.
DoorDash on What it Wanted from Automation
Observe.AI named DoorDash as one of the enterprises already running its technology. Xenia Strunnikova, Head of Customer Experience, Fraud, Trust & Safety S&O at DoorDash, said: “We didn’t want automation for its own sake. We wanted to better understand what drives customer sentiment, move beyond binary scoring, and put the customer at the center of our decisions.”
Swapnil Jain, CEO of Observe.AI, said that “AI agents are redefining how customer experience work gets done”, and that the AWS agreement is meant to extend that across the full service operation, from resolving customer requests to supporting frontline teams.
Hyperscaler Deals are Becoming the Default
Tying a CX platform to a hyperscaler is no longer unusual. Zendesk has taken a comparable route, building its AI service tooling directly on AWS, and NiCE has rebuilt its platform around agentic AI rather than adding it as a feature. Infrastructure credibility has become a sales requirement in enterprise CX, and marketplace availability shortens the buying cycle in a way that product features rarely do.
Whether any of it converts into measurable returns is a separate question. Research by CCW Europe recorded no organisation reporting enterprise-wide commercial impact from AI in its CX operations, even as 62% of CX leaders said they plan to increase AI investment over the next 12 months. Deployment infrastructure removes one obstacle. Proving the business case remains the harder part, and no cloud agreement solves it.
