Dev Dives: Build faster, test smarter, and orchestrate production-ready agents

Jul 30, 2:00 – 3:00 PM (UTC)

Event time displays in your timezone.

EMEA&APAC Virtual Events

89 RSVPs

Accelerate the build with coding agents and raise agent reliability in production. Take a tour of the latest agent enhancements: learned memory, long-horizon reasoning, external agents, evaluations, and orchestration. Learn how to apply each one and bring them together on a unified canvas in Maestro Flow.

Login to RSVP

About this event

Production agents get better when you give them the right capabilities. This session walks through new agent capabilities, when to reach for each, and how they work across the ways you build: low-code in Agent Builder, code-first with the SDK, and accelerated by coding agents. 

Learned memory lets an agent improve continuously from feedback, carrying decisions forward so accuracy compounds over time instead of resetting each run. Advanced agents take on long-horizon, large-context tasks - planning their own work, making tool calls, and spawning sub-agents to investigate open-ended problems. External agents, registered and callable over common protocols, let your automation draw on capabilities built outside your own project. Evaluations let you measure accuracy and catch regressions before an agent ships. And coding agents accelerate the build across the lifecycle, scaffolding, wiring tools, running evals, and deploying, whether you're working code-first or in Agent Builder. 

We'll show each enhancement live on the surface where it fits best, then bring everything together in Maestro Flow: agents, deterministic steps, and a human-in-the-loop checkpoint running as a single orchestrated flow. 

During this session, we'll cover: 
Learned memory: how an agent improves continuously from feedback and reuses past decisions 
Advanced agents: planning, large context, tool calls, and sub-agents for long-horizon tasks 
External agents: registering and calling an agent over common protocols as a node in your workflow 
Evaluations: measuring accuracy and catching regressions before an agent ships 
Coding agents: accelerating build, test, and deploy across Agent Builder and code-first 
Human-in-the-loop: putting a person in control before a high-stakes decision is committed 
Orchestration in Maestro Flow: agents, deterministic steps, and human approval on one unified canvas. 

 
Whether you're a low-code developer extending your first agent or an advanced developer architecting multi-agent systems, you'll leave knowing what's new and how to apply it in your own work. 

 
Speakers:

Tuan Nguyen, Principal Product Manager 

Eusebiu Jecan, Principal Product Manager 

Host:  
Cliff Simpkins 

Speakers

  • Tuan Nguyen

    UiPath

    Principal Product Manager

  • Eusebiu Jecan

    UiPath

    Principal Product Manager, UiPath

Host

  • Cliff Simpkins

    UiPath, Inc.

    Senior Director, DevRel

Organizers

  • Rohit Radhakrishnan

    UiPath

    Regional Community Manager - APAC

  • Ana-Maria Bulhac

    UiPath

    Community Associate

  • Andreea Tomescu

    UiPath

    Senior Community Manager

Explore the UiPath Community ecosystem!