ServiceNow User Group (SNUG) Miami, Florida September 29th, 2026
At the South Florida ServiceNow User Group (SNUG) meetup in Miami on September 29, 2026, Antonio Di Perna, ServiceNow Platform Owner at BankUnited joined Sanjeev Varma, Tavro AI’s Co-Founder and COO, to walk through a live deployment that is anything but a pilot. It went live in August 2026.
The session triggered a large audience engagement and discussions as many companies in attendance seem to be facing similar AI challenges, how to get started with AI at an enterprise scale and integrating ServiceNow AI with Microsoft CoPilot AI, especially if your enterprise users are working more in Microsoft Teams and CoPilot.
The takeaway from the South Florida SNUG session shared by both Antonio Di Perna and Sanjeev Varma was clear, with AI Start Small and THINK BIG!
“What’s important for employees, with all the AI agents being deployed from different platforms, is to meet them where their work is occurring and create the bridge to other AI agents and platforms needed to perform the tasks and simplify the user experience and improve productivity.”
The use case: intelligent agentic automation of IT incident management using Microsoft Copilot and ServiceNow ITSM.
The BankUnited Story: From Manual Queue to Governed AI Agent
Here is how the BankUnited deployment came together:
1. Defining the Business Problem
BankUnited’s business operations team faced three compounding pressures:
- 24×7 monitoring burden. Business user community group chats needed round-the-clock coverage for operational support. That meant human eyes on Microsoft Teams at all hours.
- High-value talent pinned to triage. Analysts with deep institutional knowledge were spending their hours on incident intake — not on the judgment-intensive work only they could do.
- Manual communication overhead. Every status update, every assignment notification, every escalation required a person to initiate it.
The solution approach was direct: deploy a Microsoft Copilot AI agent that monitors Teams, performs initial triage, creates or updates ServiceNow incidents, and manages intelligent assignment and notification — without a human in the loop for routine cases.
2. The Before-State Workflow
The before-state had six steps, all manual:
- Business user reports issue in Teams
- Team member performs initial triage
- Team member creates or updates ServiceNow incident
- Incident assigned and notification sent manually
- Resolution managed manually
- Status and communications managed manually.
3. The After-State Workflow
The after-state compressed four of those six steps into governed, automated agent actions:
- Business user reports issue in Microsoft Teams
- Agent performs initial triage
- Agent creates or updates the ServiceNow incident
- Intelligent assignment and notification
What looks simple on a slide required months of cross-functional coordination: scoping, securing business buy-in, aligning with the Microsoft 365 organization, working through reviews and approvals. The lesson Antonio Di Perna shared from the stage: “Start small, think big.” Every AI deployment touches multiple platform teams. Collaboration is not optional — it is the work.
4. Governing the Agent
The BankUnited team prepared detailed answers for Enterprise Risk Management and Information Security — not after the fact, but as part of the deployment design:
- How do you prevent the AI agent from creating unlimited incidents in ServiceNow? Rate limits, intent boundaries, and tooling constraints defined in the agent specification.
- How do you prevent the agent from accessing additional data in ServiceNow? Scoped data access, enforced at the agent tool layer.
- How do you prevent unauthorized users from triggering the agent in Copilot? Authentication and authorization controls tied to M365 identity, surfaced in the agent’s governance record.
5. What the Future Use Cases Looks Like
The BankUnited ServiceNow team is not stopping at Incident creation use case. The use cases and potentials they shared at SNUG points toward a broader agentic intelligence layer either the users are using ServiceNow Agents and/or using Microsoft CoPilot Agents integrated with ServiceNow agents.
- Predictive Intelligence: intelligent incident prioritization, similar incident lookup, intelligent incident creation, and change risk prediction
- Agent-to-Agent: service health and alerts, underutilized subscription and license detection
- MCP Server integration: connecting the agent ecosystem to Model Context Protocol infrastructure for governed, auditable inter-agent orchestration
6. SNUG Recap
Stay connected with your local South Florida ServiceNow User Group (SNUG) community, see some highlights from the SNUG event and join the community at the link South FLorida SNUG 9/29/2026 feedback, we’re you t… – ServiceNow Community