Machine Learning (ML) Engineer, Applied AI
Salary:
$140,000 - $160,000 - Per Annum
Locations:
Remote, United States
Type:
Permanent
Workplace:
Remote
Published:
September 28, 2026
Contact:
Robert Gwillim
Ref:
21710
Required Skills:
Machine Learning
Share this job
Apply
Job title: Machine Learning (ML) Engineer, Applied AI
Job type: Permanent
Salary: $140,000 – $160,000 bas + bonus
Role Location: Fully remote, US-based

The company:
Our client is an established construction technology company whose bidding platform sits at the heart of US civil infrastructure. Construction work for the vast majority of US states flows through their platform: state Departments of Transportation post projects, and contractors, subcontractors and suppliers use it to find work, bid and collaborate.

Following an acquisition two years ago, they are transforming a strong, family-built foundation into a modern technology business and investing heavily in AI. With around 300 employees (roughly half in commercial and technology roles), they are small, nimble and move at high velocity with very little red tape. Their AI roadmap is ambitious, from understanding contractor profiles and recommending relevant work, to helping contractors shape competitive strategy and find the right subcontractors to partner with.

This is an opportunity to join a small, highly capable Applied AI team as the third specialist ML Engineer, reporting to the Senior Director of AI, alongside other ML Engineers and an AI Product Manager. The team builds vertical AI and agentic capabilities used across multiple products, working directly with the product and platform engineering teams that own them. Their stack is AWS-based and makes extensive use of managed services.

Role and responsibilities:
  • Design and ship multi-step agents that use tool calling, retrieval, structured outputs, retries, fallbacks and explicit failure handling.
  • Build purpose-built, vertical AI and agentic workflows for construction bidding and procurement products, working with inputs ranging from structured bid records to long, inconsistent documents.
  • Own the eval harness: datasets, scoring rubrics, regression checks and CI integration, knowing when an LLM judge is useful and when deterministic evaluation or human review is the better choice.
  • Instrument and debug agent runs end-to-end, identifying whether failures come from the model, retrieval, tools, orchestration, source data or the interfaces between them.
  • Build ingestion and normalisation pipelines for semi-structured records and long, inconsistent documents.
  • Build MCP servers that expose product capabilities and data to AI clients through a headless interface.
  • Take full ownership of what you ship, including behaviour, reliability, latency, cost and eval coverage.
  • Work autonomously, take targeted feedback, and bring ideas and findings back to the team to help it learn and improve.

Please note: this is an applied engineering role, not a research role. You will not spend most of your time training models or operating core infrastructure.

Job requirements:
  • 2+ years building customer-facing LLM applications in production, including at least one agentic system with real tool calls where you dealt with production failure modes.
  • 2+ years of backend or full-stack engineering experience prior to that.
  • Strong Python across API, data and deployment layers.
  • Production experience with an agent orchestration framework.
  • Evaluation experience: you have built a dataset, chosen meaningful metrics and used them to catch a regression.
  • Experience with AWS or another major cloud provider.
  • Strong critical thinking and sound independent judgement, with the ability to make decisions and work through ambiguity without heavy hand-holding.
  • Thrives in fast-paced, high-intensity, low-bureaucracy environments.
  • A genuine passion for the agentic AI space and a natural drive to keep up with how quickly it is evolving.

Nice to have:
  • Supporting the development and operation of the APIs and services around AI systems.
  • Agent memory: what persists across runs and what is retrieved back into context.
  • Vector search in production, including index design, chunking strategy and retrieval quality measurement.
  • Building retrieval over long, heterogeneous documents.
  • Entity resolution or taxonomy normalisation.
  • Exposure to construction, civil infrastructure or public procurement.
  • TypeScript alongside Python.

Benefits:
  • Competitive base salary; depending on experience.
  • Fully remote working within the US, with occasional in-person team on-sites.
  • High-impact role on a small team, shipping AI features into production quickly rather than waiting months for release.
  • Ownership of production agentic systems that support the bidding and delivery of US infrastructure.
  • Direct access to senior leadership in a nimble, low-red-tape environment.
  • The chance to help shape the AI capability of a business in the middle of a major technology transformation.


Accessibility Statement: 
We make an active choice to be inclusive towards everyone every day.? Please let us know if you require any accessibility adjustments through the application or interview process.  
 
Our Commitment to Diversity, Equity, and Inclusion:
Signify’s mission is to empower every person, regardless of their background or circumstances, with an equitable chance to achieve the careers they deserve. Building a diverse future, one placement at a time.  

Check out our DE&I page here
 

Apply

We use cookies to provide the best possible experience for our users. They help us provide essential functionality and improve site performance, and allow us to offer a more personalised experience when using the site.