With six weeks left before a planned go-live, Abnormal AI made a call that ran against convention: rather than push a rollout that was already underway across the line, the company pivoted to a different platform.
Abnormal AI exists because attacks on businesses don’t look like attacks anymore. Modern email and identity threats are designed to look exactly like a normal invoice, a normal login, a normal request from a normal colleague, which is precisely what makes them hard to catch with rules-based security. Abnormal’s platform solves this by building a picture of what normal actually looks like: by establishing behavioral baselines across identities and environments correlating signals to detect anomalies, so that when something deviates, even subtly, it stands out against a backdrop the system already understands.
That approach only works if it keeps up with how fast normal changes. A model built for how a company operated last year is a poor defense against how it, and its attackers, operate today. It’s a principle Abnormal’s leadership applies well beyond the product itself: buy for where the business is headed, not just where it is now. A system that clears today’s bar but can’t flex with tomorrow’s isn’t really solving the problem.
Ashim Kapai, Head of Procurement and Accounts Payable, held Abnormal’s own procurement stack to that same standard. A platform that could get the company through go-live wasn’t enough. It needed to still make sense in two or three years, as Abnormal kept growing.
“We evaluated solutions against where we needed to be in two to three years, not just where we were at go-live,” Kapai explains. “We made the decision to pivot to Omnea because it aligned more closely with our long-term vision. It felt future-focused, intuitive for end users, and provided the level of visibility and transparency that stakeholders expect from modern procurement platforms today.”
That decision landed squarely inside the theme of this year’s DPW New York: ‘Recode.’ The conference shifted its focus from debating if AI would impact procurement, to how it would practically reshape operating models, workflows, and human-machine collaboration.
“It resonates nicely with the conference theme of Recode, because recoding doesn’t necessarily mean reworking an existing system. It’s about being willing to pivot when needed. In our case, we made the call early – before go-live – because course-correcting later costs more. It was far better than implementing something only to pull it out or replace it a year later. In my mind, recoding is really about making the right calls from a long-term perspective.”
Omnea is the AI-native platform Abnormal AI switched to. It manages supplier relationships, spend, risk and governance from a single place, giving each stakeholder, from requesters to legal, finance, information security and IT, a tailored view of exactly what they need at a given moment. Shane Gueret, Head of Implementations at Omnea, sees Abnormal’s pivot as symptomatic of a bigger shift in how procurement teams see themselves. “Traditionally, procurement has been unfairly seen as a gatekeeper,” he explains. “This is largely because it’s been hamstrung by administrative burden. We aim to champion procurement teams and give them the control and visibility they need to become a strategic asset for the business.”
For a company scaling as rapidly as Abnormal AI, that shift has significant operational implications. Being AI-native doesn’t just create opportunities, it sets a higher standard for how quickly every function, including procurement, must operate and adapt.
“That’s why speed to value is such a critical consideration,” Kapai says. “The pace of innovation and change in the market is extraordinary, and it’s a key factor in how we evaluate investments. We look for solutions that can scale with the business, accelerate outcomes, and position us to capitalize on what’s next.”
That emphasis on speed, innovation, and future-readiness surfaced again when Omnea became the first procurement platform to launch a Model Context Protocol (MCP) connector, enabling procurement teams to query supplier, contract, and spend data using natural language through AI assistants such as ChatGPT, Claude, and Microsoft Copilot, eliminating the need to navigate multiple systems or manually build reports.
For Kapai, the value of the capability was immediately apparent. “We’ve already deployed it internally and have seen tangible benefits,” he says. “What stood out to me was the pace of innovation Omnea’s team shown. Other providers we spoke with signaled a much longer timeline. My perspective was simple: by that point, the market will likely have already evolved to the next wave of AI innovation. In a market moving this quickly, you can’t afford to plan around yesterday’s roadmap.”
None of that would matter without an implementation that could keep up. Because Omnea is fully no-code, Gueret’s team can build most of the configuration themselves rather than routing it through engineering. “My team essentially acts as hybrid business analysts, solution architects, and project managers,” he explains. “We build our own custom integrations for customers, and we also act as change management consultants. We don’t outsource. We take it all on ourselves. That means we retain the knowledge and get incredibly close to our customer’s businesses and unique challenges. We spend time onsite with them, architecting and feeding back on policies.”
Inside the Omnea and Abnormal AI relationship
Six weeks from selection to go-live. That was the target Abnormal AI set after choosing Omnea, working hand in hand with Implementations Head Shane Gueret and his team to architect the project.
“We essentially started with a proof of concept. Typically, with our POCs, we want to prove out a few use cases, demonstrate that what we’ve said is technically achievable, and give customers a flavour of what Omnea can achieve.
“With Abnormal, the engagement was somewhat unique,” he explains. “The team was looking to pivot to a more modern, AI-native platform and wanted to validate not only that Omnea could meet their current operational needs, but also that it could evolve alongside the rapid pace of innovation in AI. As a result, what began as a planned six-week proof of concept quickly evolved into a full-scale implementation from day one.
“We ultimately deployed six+ distinct workflows across the company, each with its own approval structures, business requirements, and emerging AI use cases. The effort extended beyond technology implementation – it involved rethinking elements of the operating model itself. When you’re executing a transformation of that scope on such an accelerated timeline, success requires working backwards from the end state and maintaining a relentless focus on what needs to happen each day to achieve the outcome.
“It was an ambitious transformation from the outset, but our team’s mindset has always been that success is the only acceptable outcome. The six-week timeline was certainly intense, but we remained focused on execution and ultimately delivered. I believe Ashim has described it as one of the fastest enterprise implementations he’s been part of.
“That said, speed alone is not a measure of success. Many organisations can move quickly, but if they fail to establish the right foundations, the long-term outcomes often fall short. Technology implementation is only one component of a successful transformation. Equally important are executive sponsorship, cross-functional stakeholder alignment, change management, training, and governance.
“For an initiative of this scale to succeed, people need to understand not just what is changing, but why it’s changing, how it will impact them, and what success looks like. Maintaining clear communication and transparency – from end users all the way through the executive leadership team – is critical. Ultimately, sustainable transformation isn’t just about deploying a platform quickly; it’s about bringing the organisation along with you.
“Delivering a full enterprise transformation in six weeks was an ambitious undertaking, but we successfully executed against every milestone without compromising on quality, governance, or stakeholder engagement. We moved at the pace the business required while ensuring that every component of the implementation was built on a solid foundation.
“Adoption across the Abnormal teams using the new workflows has been strong. More importantly, we’ve established a scalable, future-ready procurement operating model that can evolve alongside the business and the rapid pace of innovation in AI. While we’re proud of what we’ve accomplished in a short period of time, we view this as the beginning of the journey rather than the end.”
Find out more about Omnea here.
- AI in Procurement
- Events
- Events
- Together in Events
- Together in Events