On stage were:<\/p>\n\n\n\n
- \n
- Joelle Johnson<\/a>, Lead Integrations Engineer at CYPHER Learning<\/a><\/li>\n\n\n\n
- Jason Foley<\/a>, Director of Integration Services at ClickUp<\/a><\/li>\n\n\n\n
- Shashidhar Kamath<\/a>, Director of Engineering at ORO Labs<\/a><\/li>\n<\/ul>\n\n\n\n
Here’s what led each team to that call, and where AI fits now.<\/p>\n\n\n\n
Why not keep integrations in-house?<\/strong><\/h2>\n\n\n\n
The trigger was different at each company.<\/p>\n\n\n\n
At CYPHER Learning<\/strong>, it was demand. Customers couldn’t use CYPHER without a Salesforce integration, and the company was losing money over it. Reuse mattered too: although each customer\u2019s systems were configured differently, the use cases were similar. The company ultimately chose Workato because non-technical users could build on it alongside engineers. Plus, it was reliable, the team loved its UI, and the price was right.<\/p>\n\n\n\n
At ORO Labs<\/strong>, it was a build-or-buy call. Procurement touches a long list of enterprise applications. \u201cWe had to make a call whether our engineers spend their time building integrations into all of these applications,\u201d Shashidhar explained. Ultimately, they chose to venture outside the company for a solution. \u201cWe were looking for a platform that was flexible, mature, and actually had built-in integrations,” he continued. <\/p>\n\n\n\n
At ClickUp<\/strong>, Jason’s team compared several platforms and similarly chose Workato for its flexibility and scalability. For Jason, flexibility meant managing many customers in their own environments and writing complex logic, including JavaScript libraries, inside the recipe builder. “Initially, we were starting with some customers that were pretty small, and then we got into the Fortune 100 companies,\u201d he shared. \u201cWe needed to scale from maybe a few thousand tasks to hundreds of millions of tasks in individual instances.”<\/p>\n\n\n\n
What ClickUp, CYPHER, and ORO achieved with Workato<\/strong><\/h2>\n\n\n
\n![\"The](\"https:\/\/blog2workato.wpengine.com\/wp-content\/uploads\/2026\/10\/2a641240-db3c-4e9a-9a1f-4c082d314c03-1024x683.jpg\")
Shashidhar Kamath, Joelle Johnson, Jason Foley, and Matthew Pflughoeft at WOW 2026<\/figcaption><\/figure>\n<\/div>\n\n\n Standardizing the connectivity layer paid off in how fast – and affordably – each team could deliver.<\/p>\n\n\n\n
ClickUp shipped its first integration to a paid customer 30 days after buying Workato. Repeatable templates then took an implementation that started at 60 to 80 hours down to about 10. That let Jason\u2019s team build maintenance into customer contracts and run integrations as a profitable managed service, now with many hundreds of customers.<\/p>\n\n\n\n
CYPHER got the same leverage from reuse. Joelle’s team built standard templates for the most common CRM and HRIS platforms. A change gets made once, in one place, instead of being repeated across every customer’s workspace.<\/p>\n\n\n\n
\n
“We were able to deliver integrations at speed, integrations that we trusted, integrations that our customers trusted. It was predictable.” <\/p>\n\n\n\n
\u2014 Joelle Johnson, CYPHER Learning<\/p>\n<\/blockquote>\n\n\n\n
At ORO, the biggest difference was how contained a change became. Shashidhar said that before orchestration, altering one thing meant touching every integration connected to it. Now policy changes happen in a single layer.<\/p>\n\n\n\n
AI needs context, not just access<\/strong><\/h2>\n\n\n\n
The reusable connectivity layer also gave each company a foundation for AI. But connecting systems is only the first step; an agent also needs context to interpret what it finds.<\/p>\n\n\n\n
Shashidhar used a procurement request to show the difference. An employee wants to know whether they can use a supplier, Acme, for a $300,000 consulting project. Answering that means checking the ERP, contracts, risk, and compliance, and the data is fragmented – there could be ten suppliers called Acme in the ERP and twenty accounts in the contracting system, none of them matching. Each system\u2019s AI can read its own data, but even together they probably can\u2019t answer the question.<\/p>\n\n\n\n
ORO\u2019s answer was a context layer that connects the records, the relationships between them, and the customer\u2019s policies. AI uses that context to recommend a route, such as steering the request away from Acme toward a preferred supplier. Deterministic orchestration, with Workato\u2019s integrations connecting each system, then carries it out inside the rules the customer\u2019s IT team has set. For some ORO customers, onboarding a supplier that used to take days now takes 15 minutes.<\/p>\n\n\n\n
\n
\u201cDeterministic orchestration doesn\u2019t have to be predetermined. Our AI can say, \u2018this is the best thing to do in this scenario.\u2019 But then that deterministic orchestration underlying it can make sure these things happen according to the rules IT has set up for you.\u201d <\/p>\n\n\n\n
\u2014 Shashidhar Kamath, ORO Labs<\/strong><\/p>\n<\/blockquote>\n\n\n\n
Jason described a similar idea from the integration side. A few years back, ClickUp made a bet that context had to be in the system for the AI to be worth anything. With Workato pulling in Salesforce data and customer tickets, ClickUp\u2019s AI can work across them, including through MCP servers<\/a>. In one example, it compared accounts against tickets and recommended which customers to contact first.<\/p>\n\n\n\n
Joelle\u2019s example was role context. In Salesforce, a user might look like any other user while another record lists them as someone\u2019s manager. Because CYPHER\u2019s standard Workato recipes connect those records, the AI can recognize the manager relationship and answer questions about their direct reports\u2019 enrollment and training.<\/p>\n\n\n\n
In all three cases, the AI does the reasoning, and Workato\u2019s connectivity underneath gives it connected systems to work with.<\/p>\n\n\n\n
Winning in the agentic era<\/strong><\/h2>\n\n\n\n
Asked where their AI efforts go next, each panelist described less manual work, for their own teams and for their customers:<\/p>\n\n\n\n
- \n
- ClickUp<\/strong> wants to couple orchestration with its native offerings and bring in more data, so customers can stay in one place. Its mantra has been saving users a day a week, and Jason expects AI to save more.<\/li>\n\n\n\n
- At CYPHER<\/strong>, Joelle wants integrations that set themselves up with far less manual effort, so her engineers review what’s been assembled instead of building it.<\/li>\n\n\n\n
- ORO<\/strong> is aiming for a procurement team that spends its time on strategy, with the system carrying out their decisions.<\/li>\n<\/ul>\n\n\n\n
\n
\u201cWe were building an Iron Man suit with superpowers for the procurement department, through orchestration. Then along came AI. Now we can put Jarvis inside this Iron Man suit and automate a lot of this stuff. But not everything – we still have someone in the suit, overseeing things.\u201d <\/p>\n\n\n\n
- At CYPHER<\/strong>, Joelle wants integrations that set themselves up with far less manual effort, so her engineers review what’s been assembled instead of building it.<\/li>\n\n\n\n
- Jason Foley<\/a>, Director of Integration Services at ClickUp<\/a><\/li>\n\n\n\n