More than half of the Fortune 500 relies on Workday for HR, finance, and IT. In 2025, Workday's leadership team saw the market opportunity for building agentic solutions and set an ambitious goal for agents in production by the end of 2026. Deploying autonomous software within such heavily regulated domains typically carries massive engineering overhead, as internal services must strictly satisfy corporate audit, compliance, and data retention policies.
The initial execution fell on an Agent Platform team consisting of just four engineers. The central question facing the organization was whether such a small unit could deliver a scalable agent runtime capable of meeting Fortune 500 compliance without triggering a massive expansion in engineering headcount. Workday proved that standardizing on existing orchestration frameworks rather than building custom backend architecture allowed this four-person team to support the enterprise's functional demands.
Standardizing the Orchestration Layer
To manage this infrastructure footprint, the team avoided building a bespoke backend engine from scratch. Instead, Workday leveraged LangGraph for agent orchestration and LangSmith for tracing, observability, and evals. Choosing LangGraph gave internal developers flexibility at the design layer, as its graph-based model enabled engineers to assemble everything from simple ReAct loops to complex deterministic workflows without the core platform prejudging what an agent should look like.
The small team also built an internal Python SDK that Alex Gatto, Senior Machine Learning Engineer, described as "Workday LangChain." This library wrapped LangChain primitives and routes them through Workday's internal services, including an LLM gateway, a RAG retrieval service, and other internal tooling. As Gatto noted, the infrastructure offloaded standard backend hurdles directly onto existing tooling:
"One of my favorite things about LangChain’s Agent Server and LangSmith is just how much you get for free. Just because of how much they've thought about agent development."
Alex Gatto emphasized that "the backend isn't the problem," and reported that "LangChain took care of it for us." This abstraction established a standardized layer across internal tools that had previously operated in isolation.
Governance and Compliance Under Audit Constraints
The most pressing technical objection to this lightweight structure was whether a four-person squad could guarantee rigorous governance and enterprise security. The team addressed this upfront by releasing Workday's Agent System of Record as their first major deliverable. This governance layer controls which APIs the agents are allowed to access, running directly on top of the backend server and eval platform.
Because LangChain's agent server implementations are built with A2A and MCP support, every agent server the Agent Platform team ran was already compliant with both protocols. Whenever internal developers modify a graph, those updates were immediately reflected in the server environment without requiring secondary backend engineering.
"There's just no way that we could have delivered value to our customers without LangSmith’s speed and maturity."
According to Alex Gatto, Senior Machine Learning Engineer, this foundation allowed the team to immediately answer that strategy for A2A and MCP protocols was already in production when senior leadership asked about it.
Compressing Development Cycles
Historically, building a prototype agent took 4 to 5 months, with no guarantee it could ever reach production. Under the new platform architecture, engineering teams can have a prototype running in a staging environment in under a day, with authentication and security requirements baked in.
The operational shift translated directly into organizational growth. After establishing the core platform, the initial group of four has now grown to roughly 30 members, with continued investment planned.
Alex Gatto confirmed that because of the success of their partnership with LangChain and what they were able to accomplish, they have more than tripled the size of their team.
What it signals
Engineering payroll for sprawling backend teams becomes entirely unnecessary when launching complex enterprise software requires merely four specialists. Workday demonstrated this directly: instead of writing a proprietary server engine, the team assembled their production infrastructure around ready-made LangChain components, bypassing traditional architectural bottlenecks. According to the company, relying on LangSmith and LangGraph handled orchestration and validation out of the box while satisfying enterprise compliance across Fortune 500 standards. Standardizing agent orchestration around turnkey frameworks reduces platform engineering from a massive headcount sink into a lean internal unit.