Workato

We turned on Enterprise MCP servers for our go-to-market team last fall. In our last quarter, that drove an additional $1M in revenue not from a new product or pricing change. Sellers spent more time selling because LLMs could pull and update data (take ACTION!!) they needed in enterprise systems. Plus, managers had more accurate data and visibility into deals that they could pull into the quarter. That's what happens when your LLMs can actually access the systems that run your business. I shared these numbers last night at our Enterprise MCP Happy Hour at Workato's AI Lab in San Francisco. Full room of CIOs and IT leaders all wrestling with the same question: How do get true business impact with AI? MCP is the foundation. But without orchestration, governance, and trust built in from the start, you're just building chatbots. Our journey to get to true business outcomes with AI: OpenAI agreement in late 2023. First agents in 2024 with low adoption. ChatGPT and Claude for everyone in 2025. First MCP server built last summer. Each step built on the last but adoption exploded once we connected LLMs to real enterprise data through MCP. The energy in the room confirmed what I keep hearing from peers. CIOs are ready to lead this. We just need to start building. Thanks to everyone who came out last night.

Wow! More than 100 customers, partners and friends of Workato here at the SF AI Lab today. We are having a conversation about agentic orchestration and enterprise MCP with some of the hottest companies in the world. Exciting to see the leadership and innovation in this room. Thanks for welcoming us into the AI community, San Francisco! OpenAI CoreWeave Scale AI Amazon Web Services (AWS) Perplexity and 95 more The mini grilled cheese and tomato soup bites were awesome, too

Most people in the room had built an agent in the last 30 days. Almost none had shipped one. That gap isn't a technical problem. It's three problems layered on top of each other: architecture, evaluation, and absorption. And on one Tuesday night in Dogpatch, three practitioners mapped all three without knowing they were building the same framework from different directions. A Workato Staff AI Engineer explained why single agents fail before you finish celebrating the demo. A Databricks advocate named what nobody warned software developers about: prompt engineering under eval pressure is gradient descent. A Google engineer proved that the hardest part of agent adoption had nothing to do with the model. And an audience question about DAGs pointed quietly toward where multi-agent architecture actually needs to go next. Here's what makes Workato's AI Hub different from a meetup with good snacks. Every question asked in that room is unfiltered practitioner intelligence. Live signal about where developers are actually stuck. Where frameworks are failing. Where the next adoption friction is forming. The organizations that feed that signal back into what they build next are running a feedback loop their competitors cannot manufacture. Workato is listening. The eval loop applied to the organization itself. What gets built from evenings like this will separate demonstration from deployment. Field report from the Workato AI Hub in Dogpatch:

🎉 What a week! Congratulations to our newest batch of innovators who completed our inaugural Workato 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻𝘀 𝗜 + 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗪𝗼𝗿𝗸𝘀𝗵𝗼𝗽𝘀 across #Sydney and #Canberra! 👏 In Sydney, this group went from building their first recipes to getting hands-on with agentic AI and discovering what's possible with enterprise MCP. We wrapped the week in Canberra, where discussions ran so deep that even though we planned to finish early, almost everyone stayed back. The calibre of questions blew us away. Practitioners from 𝗴𝗼𝘃𝗲𝗿𝗻𝗺𝗲𝗻𝘁, 𝗳𝗶𝗻𝗮𝗻𝗰𝗶𝗮𝗹 𝘀𝗲𝗿𝘃𝗶𝗰𝗲𝘀, 𝗿𝗲𝗮𝗹 𝗲𝘀𝘁𝗮𝘁𝗲, 𝗰𝗼𝗻𝘀𝘁𝗿𝘂𝗰𝘁𝗶𝗼𝗻, 𝗵𝗼𝘀𝗽𝗶𝘁𝗮𝗹𝗶𝘁𝘆, 𝘁𝗲𝗰𝗵, 𝗺𝗲𝗱𝗶𝗮, 𝘀𝘂𝘀𝘁𝗮𝗶𝗻𝗮𝗯𝗶𝗹𝗶𝘁𝘆, 𝗰𝗼𝗻𝘀𝘂𝗹𝘁𝗶𝗻𝗴 and more, all reimagining how work gets done. In their own words: Shirley Zhao: "𝘼𝙄 𝙞𝙨 𝙩𝙝𝙚 𝙗𝙧𝙖𝙞𝙣, 𝙈𝘾𝙋 𝙞𝙨 𝙩𝙝𝙚 𝙖𝙧𝙢, 𝙨𝙠𝙞𝙡𝙡𝙨 𝙖𝙧𝙚 𝙩𝙝𝙚 𝙩𝙤𝙤𝙡𝙨." Matthew Adisusila: "𝙒𝙤𝙧𝙠𝙖𝙩𝙤 𝙝𝙖𝙨 𝙧𝙚𝙫𝙤𝙡𝙪𝙩𝙞𝙤𝙣𝙞𝙨𝙚𝙙 𝙞𝙣𝙩𝙚𝙜𝙧𝙖𝙩𝙞𝙤𝙣 𝙨𝙤 𝙨𝙚𝙖𝙢𝙡𝙚𝙨𝙨𝙡𝙮 𝙩𝙝𝙖𝙩 𝙙𝙚𝙨𝙞𝙜𝙣𝙚𝙧𝙨 𝙖𝙣𝙙 𝙙𝙚𝙫𝙚𝙡𝙤𝙥𝙚𝙧𝙨 𝙘𝙖𝙣 𝙛𝙤𝙘𝙪𝙨 𝙢𝙤𝙧𝙚 𝙤𝙣 𝙧𝙚𝙨𝙤𝙡𝙫𝙞𝙣𝙜 𝙗𝙪𝙨𝙞𝙣𝙚𝙨𝙨 𝙘𝙖𝙨𝙚𝙨 𝙖𝙣𝙙 𝙖𝙪𝙩𝙤𝙢𝙖𝙩𝙞𝙣𝙜 𝙚𝙫𝙚𝙧𝙮𝙙𝙖𝙮 𝙩𝙖𝙨𝙠𝙨." Iain T.: "𝙒𝙤𝙧𝙠𝙖𝙩𝙤 𝙞𝙨 𝙖 𝙢����𝙨𝙩 𝙛𝙤𝙧 𝙖𝙣𝙮 𝙤𝙧𝙜𝙖𝙣𝙞𝙨𝙖𝙩𝙞𝙤𝗻." Tim Higgins: "𝙄𝙣𝙨𝙞𝙜𝙝𝙩𝙛𝙪𝙡 𝙨𝙚𝙨𝙨𝙞𝙤𝙣 𝙬𝙞𝙩𝙝 𝙝𝙞𝙜𝙝𝙡𝙮 𝙘𝙤𝙢𝙥𝙚𝙩𝙚𝙣𝙩 𝙙𝙚𝙡𝙞𝙫𝙚𝙧𝙮." Andy Foo: "𝘽𝙚𝙨𝙩 𝙖𝙥𝙥𝙡𝙞𝙘𝙖𝙩𝙞𝙤𝙣 𝙖𝙣𝙙 𝙗𝙚𝙨𝙩 𝙞𝙣𝙨𝙩𝙧𝙪𝙘𝙩𝙤𝙧." Roger Bush: "𝙂𝙧𝙚𝙖𝙩 𝙬𝙖𝙮 𝙩𝙤 𝙪𝙣𝙙𝙚𝙧𝙨𝙩𝙖𝙣𝙙 𝙩𝙝𝙚 𝙥𝙧𝙤𝙙𝙪𝙘𝙩 𝙖𝙣𝙙 𝙜𝙚𝙩 𝙚𝙭𝙘𝙞𝙩𝙚𝙙 𝙖𝙗𝙤𝙪𝙩 𝙬𝙝𝙖𝙩 𝙞𝙩 𝙘𝙤𝙪𝙡𝙙 𝙙𝙤 𝙛𝙤𝙧 𝙤𝙪𝙧 𝙗𝙪𝙨𝙞𝙣𝙚𝙨𝙨." Neil Saunders: "𝘼 𝙨𝙡𝙞𝙘𝙠, 𝙚𝙭𝙥𝙚𝙧𝙩𝙡𝙮 𝙤𝙧𝙜𝙖𝙣𝙞𝙨𝙚𝙙 𝙚𝙫𝙚𝙣𝙩 𝙬𝙞𝙩𝙝 𝙖 𝙛𝙪𝙡𝙡𝙮 𝙥𝙧𝙚𝙥𝙖𝙧𝙚𝙙 𝙚𝙣𝙫𝙞𝙧𝙤𝙣𝙢𝙚𝙣𝙩, 𝙩𝙝𝙤𝙪𝙜𝙝𝙩𝙛𝙪𝙡 𝙢𝙖𝙩𝙚𝙧𝙞𝙖𝙡𝙨, 𝙖𝙣𝙙 𝙖𝙣 𝙚𝙭𝙘𝙚𝙥𝙩𝙞𝙤𝙣𝙖𝙡 𝙩𝙧𝙖𝙞𝙣𝙚𝙧." Ideas are already flowing — enterprise-grade Genie rollouts in finance, service desk automation, MCP skills for agents, and streamlining manual processes across government, hospitality, and beyond. Massive thanks to Deloitte for sponsoring their stunning Sydney venue at Quay Quarter Tower. Thank you Sasindran Sivapalan, Prachi Singh, and David Johnsson for making it happen, and Santosh Jungari, Ravi Timmavajjula, Jieshun Wang, Ricardo Brillembourg, and Elsayed Ahmed for training alongside our community. Huge shoutout to Mary who delivered every session flawlessly across 3 cities this fortnight. And to our ANZ team — Joy Botha (She / her / hers), Jason Wills Chiu, Jack Gibson, Cameron Kahler, Kokil Handa, Luis, and James Hancock 🙏 To all our learners: You've been so energising to engage, and leave us so inspired with the ideas you're brimming with! Let's turn them into impact. 💡 🎬 Here's a video to sum it up! #Workato #AgenticAI #MCP #EnterpriseOrchestration #AgenticOrchestration #Automation #Deloitte