July DLT Monthly Meeting Summary - "IT Automation"
For the July DLT monthly discussion, we focused on the topic of IT Automation. We had a rich conversation about how emerging technologies are creating opportunities and new challenges to consider while pursuing such initiatives.
Approaches to IT Automation
Customer service and back office: We started with members in the financial services industry discussing opportunities at their company in areas such as call center/back-office operations, especially customer service processes (such as deceased account workflows), as well as using agentic AI to enhance agent knowledge and efficiency. Another member shared the use of agentic AI for exception-based mortgage underwriting.
Regulatory: Another member mentioned using AI to parse and manage state-by-state DUI program rules for breathalyzer devices. A human-in-the-loop model ensures accuracy while reducing legal and customer service errors.
Corporate Functions: One member shared their focus on ITSM, HR, Finance, and eventually core operations - using agentic AI. Their intent is to be able to drive productivity gains and reduce costs as they scale.
R&D: Another member shared how their AI agent lives in R&D: and turns research outputs into structured reports.
One member shared the following construct for how they are pursuing different opportunities:
Communicative (e.g., product descriptions, SEO, HR branding)
Collaborative (e.g., content translation, personalization)
Operative (longer-term focus on integrated systems and governance)
Challenges & Governance Approaches
Balancing speed and risk: The members shared a number of issues such as organizational silos and slow governance pathways.
Working groups: A number of members shared that they are forming AI working groups with legal, risk, compliance, and tech.
Integrating in existing systems: Some were embedding AI within existing governance channels.
Agent Orchestration: Establishing platforms for agent orchestration (e.g., A2A protocols).
Data tagging: Data tagging and cost tracking through FinOps teams have proven important.
Change Management Practices
Business leadership: While tech and risk teams are often leading AI adoption, that won't work as well as when business leadership is also involved to help scale the opportunity more broadly. Leadership can be involved in steering committee meetings.
Culture and training: There needs to be a cultural shift and additional workforce retraining to reduce resistance and fear.
Ongoing events and awareness: Initiatives like quarterly AI days help raise awareness and excitement company-wide.
Virtual employees: Tracking “virtual employees” like regular agents in Workday might be one way to integrate into existing systems and practices.
Additional Insights
Measurement and metrics: Members used a number of measurements such as cost per transaction or file; efficiency ratios (esp in finance); operational throughput and scaling; time-to-value and speed of deployment.
Experimenting: Some companies set up internal experimentation frameworks (e.g., “company.ai” platform). Others leverage partnerships with cloud providers like AWS or Google to subsidize early experimentation.
Solution Providers: Some vendors mentioned included Supervity, Levity, and HappyRobot. Each has different prebuilt AI agents. Additional solutions mentioned were Copilot, Writer, Pixi.
We want to thank all who participated and shared their approaches. We all get smarter when we can collaborate and share in this way.
