Agentic AI Engineer

wsb•Austin, TX

1d•Onsite

About The Position

Forge ahead with WSB. Our Application Development team is seeking an Agentic AI Engineer. This role builds AI-powered tools, integrations, and agents that ship to 1,000+ internal engineers and technical staff, and increasingly to our commercial product offerings. The work connects systems, reduces manual effort, and creates real leverage for people who design, build, and maintain infrastructure across the country. The Agentic AI Engineer will approach problems with a first principles mindset regardless of experience level. We believe ‘expert’ thinking blocks discovery; curiosity and open-mindedness drive it. The successful candidate is relentless about solving problems, an effective listener, creative in their thinking, and able to communicate technical topics cleanly to audiences of varying technical expertise. They can go from 0-100% understanding of a new technology when a solution requires it, and knows how to engage other technical teams for independent review. The successful candidate will be mentored by the Lead Solutions Engineers and Architects, and the broader Application Development and Information Technology teams. This person will be working in heavy collaboration with technical specialists in various internal systems and disciplines, as we continue to evolve and scale our internal IT Partner Services. We apply the scientific method to our work: propose, test, measure, make decisions based on evidence, make our assumptions explicit, and re-test previous findings as the landscape changes. Staying technical does not mean staying put. Engineers who remain at the front of their craft advance and are developed on that basis. A management track isn't required. Technical leadership is.

Requirements

Scripting and general-purpose programming: PowerShell and Python; comfort with at least one and willingness to learn the other is expected.
Experience working with LLM APIs (Claude, Copilot, or similar).
Prompt engineering, retrieval-augmented generation (RAG), function/tool calling, and building skills or agents that compose into larger workflows.

Nice To Haves

REST API development.
Version control with Git.
Strong understanding of programming concepts: data structures, logic flow, iteration, functions.
Strong understanding of software architecture and computing fundamentals. Agentic tools like Claude and code quality tools like SonarQube accelerate implementation and review, but they do not replace the planning, specification, architectural judgment, and opportunity-cost evaluation that separates good solutions from expensive ones. Success in this role comes from knowing what to build and why, then using modern tools to build it well.
Strong understanding of information security, privacy best practices, and document control for data systems handling confidential or operationally critical information.
MCP (Model Context Protocol) experience or strong curiosity about it. This is an essential and constantly evolving part of our stack.
Knowledge graph systems, SQL, or non-traditional data stores.
Familiarity with agentic patterns: systems that reason step-by-step, call tools and observe results, critique and revise their own output, or coordinate across multiple specialized agents.
Experience with LLM evaluation, observability, or prompt testing frameworks.
Prior work in the AEC, infrastructure, or engineering consulting space.
Experience building software and solutions that ship to real end users (internal or commercial).

Responsibilities

Design and build AI agents and integrations using modern LLM platforms and frameworks.
Connect internal and partner systems (Azure, Microsoft 365, SharePoint, Freshservice, Bentley MicroStation / OpenRoads / ProjectWise, Autodesk platforms, GIS, and other vendor tools) to LLM-powered workflows that deliver measurable business value. This list is a subset; our reach and scope are vast.
Build systems that reason through problems, call tools, iterate on their own output, and coordinate with other agents.
Research and test emerging technologies. Document the costs, risks, and tradeoffs of each approach so our Information Technology and Application Development teams can make informed decisions.
Ship software that runs in production and affects how engineers, project managers, and leadership do their jobs every day.
Work on both internal tools and commercial product lines as WSB expands its software offerings to external clients.
Collaborate with engineers, project managers, legal, finance, business unit leaders, and technology vendors and partners. We currently work closely with Bentley, Esri, 4M Analytics, HeadLight, and others, and the list continues to grow.