About this role
Summary: Build production AI agent pipelines, tool integrations, and features for an agentic platform. Own features end-to-end from prototype to production, ensuring reliability and correctness through rigorous testing and monitoring.
POSITIONSUMMARY:
We
are looking for an AI Engineer to be a core builder on our client’s agentic
platform. You will build the agent pipelines, tool integrations, and production
features that make AI agents do real, reliable work across the company —
working closely with senior engineers who set the patterns you build against.
This
is a hands-on role for an engineer who wants to build production AI systems,
not just demos. You will own features end to end from a working prototype to a
component running against real company data and workflows. You will learn the
failure modes of AI systems by operating them, get your code reviewed at depth,
and steadily take on harder and harder parts of the platform.
This
is not a role at a software company, and it is not a role where you write a
little code on the side. It is a role helping build the foundation an entire
organization will run AI on for the next decade. The portfolio is messy, the
problems are real, and the path to grow is clear.
KEY JOBRESPONSIBILITIES:
Build agent pipelines and integrations. Implement the
working pieces of our agentic systems — agent pipelines, the tool and data
integrations that connect agents to Autodesk, SAP, Databricks, and other
systems, and the paths agents call to get real work done.
Ship to production and keep your systems healthy. Take your work
from prototype to production: deployment, monitoring, controlled release, and
fixing what breaks. The bar is not a demo that runs once — it is a component
running reliably against real workflows, with you accountable for the parts you
ship.
Write tests and evaluation checks. Verification is
what separates a demo from production. You will write the tests, contract
checks, and evaluation cases that prove your agents produce correct output
before it ships — and the regression checks that catch it when their behavior
drifts.
Build against the standard — and help improve it. You work inside
patterns set by senior engineers. Building well against them is most of the
job; the other part is flagging honestly, with a concrete example, when a
pattern doesn’t fit what you’re building. That feedback is how the platform
gets better.
Grow into the harder work. This is a growth role. You will pair
with senior engineers and steadily take on more of the core platform — a harder
integration, an evaluation gate, a piece of the orchestration loop.
REQUIREDKNOWLEDGE, EDUCATION, EXPERIENCE, SKILLS, AND ABILITIES:
- 4–6 years in software engineering, with a track record of
shipping features to production and owning them afterward
- Hands-on experience building with LLMs or AI agents — you
have built something real with an agent framework, an LLM API, or a
tool-calling system, beyond using AI in your own workflow
- Strong software engineering fundamentals — testing, version
control, code review, debugging — and the discipline to apply them even when
the output is nondeterministic
- Comfort in a cloud environment (AWS preferred): you can
deploy, monitor, and troubleshoot a service you built
- Some data or integration experience: calling SaaS APIs,
moving data between systems, or working against a data warehouse or lakehouse
- A builder’s instinct: you ship the working thing, iterate in
production, and ask for help before spinning for a day
NICE TO HAVE:
- Hands-on experience with an agent framework or SDK
- Experience writing tests or evaluation cases for AI or LLM
output
- Experience with Databricks or a similar lakehouse platform
- Experience building or consuming tool-integration layers
(e.g., MCP servers) between AI systems and enterprise applications
- Experience in a non-software-company engineering organization
— internal tools, corporate IT, or similar
- Public work — open source, a side project, anything that
shows how you build
WHY THISROLE:
You
will build production AI systems at the moment an organization decides to
invest seriously in them — with senior engineers and an architect as people you
learn from, not gatekeepers who block you. The platform is early, and that is
the appeal. You are not maintaining someone else’s infrastructure; you are
helping build the platform the company will run on for the next decade, and
growing into owning the hardest parts of it