UiPath (NYSE:PATH) outlined its product strategy and roadmap during a virtual fireside chat and product strategy overview featuring Founder and CEO Daniel Dines and Chief Product and Technology Officer Raghu Malpani. The event focused on how UiPath is positioning its automation platform for what executives described as a shift from traditional RPA toward agentic automation and “business orchestration,” emphasizing deep enterprise integration, governance, and the use of coding agents across the automation lifecycle.
Leadership discussion: AI, RPA, and UiPath’s role
Dines and Malpani opened with a discussion on why Malpani joined UiPath after roles at Microsoft and Facebook, with Malpani citing cultural alignment and what he described as a timely transition “from RPA to agentic automation to business orchestration.” Dines credited Malpani as “the architect behind our big push into Maestro,” which he described as a modern workflow engine underpinning the company’s direction.
Dines said he is “most excited about” the emergence of coding agents, describing them as a way to build automations at larger scale. He said UiPath has “pivoted the entire company” to enable its platform to be used “primarily” by coding agents, adding that UiPath intends to be “coding agents agnostic,” supporting tools such as Claude Code and Codex. Dines framed the goal as enabling AI-assisted automation across the full lifecycle: requirements gathering, specification, architecture, building artifacts (including RPA, Maestro, and Document Understanding), testing, deployment, monitoring, and exception handling in production.
Product strategy: an “agentic business orchestration platform”
Presenting UiPath’s roadmap, Malpani pointed to what he described as enterprise constraints, including integration challenges and governance concerns. He cited a recent CIO survey and said nearly half of CIOs identified connecting AI agents to existing systems—databases and CRMs—as a major barrier, alongside data quality issues. He also highlighted “shadow AI,” noting that “about 20-odd percent of deployments are already unauthorized,” arguing this increases demand for unified platforms combining integration and governance.
Malpani described UiPath’s “agentic business orchestration platform” as moving up the value chain:
- Task automation: UI automation and enterprise connectivity across applications and interfaces.
- Process orchestration: Maestro coordinating robots, humans, and AI agents, including long-running workflows that can span “hours, days, weeks, even months.”
- Agentic case management: A new “case manager agent” designed to triage and resolve dynamic, exception-heavy processes.
- Vertical solutions: Industry- and department-specific offerings aimed at delivering end outcomes.
Malpani said UiPath is embedding coding agents “natively” into its platform to accelerate every phase of building, operating, and governing automations. He said coding agents can capture requirements from documents, create production-ready workflows “with guardrails,” diagnose errors via logs, propose fixes, and support operations such as managing machines and meeting SLAs. He added that UiPath’s low-code experiences remain important, particularly for users who want to visually verify automation intent.
Maestro orchestration and case management
Malpani described Maestro as a control-tower layer that is “endpoint agnostic,” able to orchestrate systems of record and agents built outside UiPath. He emphasized “durable execution,” saying the platform must recover from failures and resume work “exactly where it left off,” supported by an “event-sourced engine.” He also said Maestro provides auditing, governance, and human-in-the-loop checkpoints.
To illustrate complexity, Malpani walked through an example insurance claims process involving parallel steps, escalations to humans, AI agents, deterministic automations, and APIs. He argued simpler workflow tools can handle a “happy path,” but break down under exceptions and handoffs, which is where orchestration becomes critical.
He said UiPath’s “case manager agent,” described as launching in May, is designed to maintain state and context across stages of a case and dynamically decide paths rather than rely only on a fixed flowchart.
Vertical solution demo: loan origination QA/QC
Mark Rubinstein, Director of Product Management for Financial Services Solutions, demonstrated what he called the UiPath Solution for Loan Origination. Rubinstein said UiPath spent months with “dozens of lenders” observing loan origination work across loan officers, processors, underwriters, and QA analysts. He described fragmented systems and manual verification as drivers of delays, higher costs, and defects.
Rubinstein cited several metrics to describe the problem space, including “42 days on average to close a conventional mortgage,” “nearly $11,000 to originate a single loan,” and “47%” of critical defects tied to “manual verification and calculation.” He said the solution includes two modules: loan setup (between application and underwriting) and a QA/QC module (after underwriting and post-closing), designed to connect to existing loan origination systems, core banking, and content management without “rip and replace.”
In the demo, Rubinstein showed a QA/QC workflow that aggregates data and documents into a unified view, automates document checks using agents, deterministic workflows, and document extraction, and allows human reviewers to confirm results for audit purposes or override agent findings with notes. He also demonstrated automated issue summaries for escalation emails and automated report generation for audits. Rubinstein said the application was built as a UiPath Process App on top of Maestro case management.
Customer example: One New Zealand on orchestration and speed
UiPath also featured a customer discussion with Jason Paris, CEO of One New Zealand. Paris said the company has deployed AI for over a decade, including “thousands of RPAs,” and aims to be “the most AI-enabled telecommunications company on the planet.” He described the company’s approach as “AI first, but human where it matters,” and said its organization is broadly “process mapped, rewired, automated, and having agentic tools layered on top of it.”
Paris described Maestro as an “orchestrator over the top of our AI and systems and people,” particularly valuable in a legacy environment with multiple stacks and billing platforms. He said One New Zealand did not need “a major re-platform or replacement” to deploy UiPath. As one example, he described a business customer handset replacement process that previously took “4-5 days” end-to-end, which One New Zealand reduced to “5-10 minutes” using process mapping, automation, and orchestration over existing workflows.
Paris also discussed deployment speed, saying the company built its “very first agentic agent within about 12 hours,” followed by several weeks to deploy after cleaning data and testing. He attributed speed to both the technology stack and a “two-in-a-box model” pairing company experts with UiPath experts.
In closing remarks during the Q&A, Dines differentiated “agent-to-agent orchestration” from “process orchestration,” arguing enterprises prefer injecting AI steps into deterministic orchestration engines to meet regulatory, governance, and audit needs. He concluded that UiPath is focused on bringing coding agents into the platform, calling it an accelerator for adoption, and reiterated the company’s emphasis on a modern workflow engine, BPMN-based process description, and task automation connectivity across legacy and modern systems under integrated security and governance.
About UiPath (NYSE:PATH)
UiPath Inc provides an end-to-end automation platform that offers a range of robotic process automation (RPA) solutions primarily in the United States, Romania, the United Kingdom, the Netherlands, and internationally. The company offers a suite of interrelated software to build, manage, run, engage, measure, and govern automation within the organization. Its platform's embedded AI, ML, and NLP capabilities improve decisioning and information processing; emulate human behavior allows organizations to address a myriad of use cases; emulate human behavior allows organizations to address a myriad of use cases; multi-tenant platform enterprise deployment with security and governance and Automation Cloud, which enables customers to begin automating without the need to provision infrastructure, install applications, or perform additional configurations; intuitive interface and low-code, drag-and-drop functionality; signed to enable people and automations to work together; and tracks, measures, and forecasts the performance of automations, enables customers to gain powerful insights and generate key performance indicators with actionable metric.
