Identity and Delivery in the Age of Agents | Harness X Okta
AI agents are doing more than generating code or answering questions. They can access systems, interact with applications, trigger workflows, and take actions across the software delivery lifecycle.
As agents take on more responsibility, engineering and security teams need to think differently about identity, access and control. That means establishing who or what an agent is, defining what it can access and do, and maintaining visibility into its actions as agents become part of production applications and engineering workflows.
Harness and Okta are bringing together engineers, security practitioners, and technology leaders to look at this shift from two connected perspectives: identity and software delivery.
Expect practical conversations on securing access for AI agents, governing their actions, and maintaining control as agents become more deeply integrated into how software is built and delivered.
The meet-up will feature three talks followed by an audience Q&A, with time afterwards to connect with speakers and fellow attendees
Speakers and session topics will be announced soon.
💻 Bring your laptop
This is a hands-on meet-up, so you won’t just be listening to the sessions. You’ll get to follow along with the speakers and try things out for yourself. Bring your laptop fully charged and ready to go.
Agenda
11:00 – 11:30 AM — Registration & Networking
11:30 – 11:40 AM — Welcome: Okta × Harness
11:40 AM – 12:10 PM — Cross-App Access for AI Agents by Vanshika Pandey, Senior Builder Advocate, Okta: A practical introduction to OAuth, OIDC and XAA, followed by a hands-on walkthrough of how identity and authorization work when AI agents need to access applications and services.
12:10 – 12:40 PM — Securing the Machine: Architecting Access for AI Workers and MCPs by Shashank Singh, Manager, Software Engineering, Harness: As AI agents and MCP clients begin taking actions across software delivery, traditional access models need to evolve. This session looks at fine-grained permissions, authentication, and authorisation for giving agents the access they need without giving them more than they should have.
12:40 – 1:10 PM — Session 3: Speaker and topic to be announced
1:10 – 1:40 PM — Audience Q&A
1:40 – 3:00 PM — Networking & Refreshments
What You’ll Walk Away With
A better understanding of how identity changes when AI agents can access systems and take actions
Perspectives on managing access and control as agents become part of engineering workflows
Practical considerations for bringing agents into production environments responsibly
A chance to exchange experiences with engineers, security practitioners and technology leaders working through similar problems
Who Should Attend
Software Engineers working with AI-enabled applications and workflows
Platform, DevOps and SRE teams
Security and Identity Architects
Engineering Managers and Technical Leads
Practitioners responsible for bringing AI agents into production
Spots are limited and registrations are reviewed. Apply if you’re actively building, securing or operating in this space.