# Sr AI Engineer (Emerging AI) — DAT

- Location: Seattle, WA
- Posted: Oct 12, 2026
- Status: open (checked Nov 21, 2026)
- Apply (employer's site): https://careers.dat.com/jobs/6206789004?gh_jid=6206789004
- Employer record: https://visajobs.io/company/dat-solutions.md

## Sponsorship history
- Tier: Good history
- Occupation (SOC): Software Developers (15-1252)
- Employer's certified LCAs for this occupation in WA, FY2024–FY2026: 3
- …in the whole US: 6
- With this exact title: 0
- Median DOL prevailing wage, level I / II: $111k / $137k

## Description
About DAT
DAT Freight & Analytics is an award-winning employer of choice and a next-generation SaaS technology company that has been at the leading edge of freight and logistics innovation for nearly five decades. Founded in 1978, DAT operates the largest freight marketplace in North America — processing 250 million+ load posts annually and maintaining one of the largest repositories of freight market transaction data in the world. On a defined path to $1 billion in revenue, DAT deploys a suite of software solutions, machine learning models, and intelligent automation tools that help brokers, carriers, and shippers price freight accurately, source capacity, reduce risk, and operate more efficiently. With nearly 700 teammates across offices in Denver, CO; Portland, OR; Seattle, WA; Springfield, MO; Toronto, ON; and Bangalore, India, DAT combines the credibility of a multi-decade market leader with the drive of a company that is not done disrupting the industry it helped build. For more information, visit www.DAT.com
Job Application Deadline: 01/31/2027
The Opportunity
As a Sr AI Engineer, you're an integral part of the team who will build and own the platform that power our agentic solutions at DAT. Freight is an uncertain business, and the models we ship reduce that uncertainty: rate forecasts, load-to-truck matching, document extraction, fraud signals, and the agentic workflows our brokers and carriers use to move freight faster. None of that reaches a customer without a platform that makes training, serving, evaluating, and monitoring models routine instead of heroic.
You'll help build the platform, from the model gateway and inference layer through feature and vector storage, evaluation harnesses, production observability, agentic framework (A2A) all on a spine based on secure access and utilization. This is a highly visible role for an engineer who wants their work multiplied across every AI team in the company.
What You'll Do
Build and Ship: Write production code for the agentic services and platform components that power DAT's agentic mesh. This powers the systems our product and engineering counterparts build from.
Deliver Against Specification: Build an approved specification which includes the problem, the non-goals, the non-functional budget, the trust boundary it crosses and deliver working software against it. Flag drift between what was asked for and what you were given before you write the first line, not after.
Own Your Tests: Write the test suite for what you build including unit, integration, and evals from the specification itself, not from your own implementation, so a green suite means the requirement is met, not that the code agrees with itself. Your tests and observability are what raise, or cap, the repository's autonomy ceiling.
Build Agentic Systems: Build and operate agents, tools, and multi-agent workflows on DAT's agent mesh which includes tool-calling, retrieval, memory, and orchestration that run inside the entitlements resolved for the request and never reach past them.
AI Infrastructure on AWS: Build and operate production AI infrastructure on AWS — Bedrock,Agent Runtime Core, Strands, Lambda, S3, etc
Design: You don't own the architecture, but you do own designs built from the premise. Bring informed pushback to decisions made by the team, flag what won't hold up in production, and show up to the conversation with an opinion, not just an implementation.
Author Designs: Write the high-level design (HLD) for what you build. The design, the tradeoffs you made, and why so the next engineer, and the next team, can pick it up without you in the room.
Review for Fidelity: Review AI-generated and human-written changes for the systems you own, holding the line on fidelity to the spec rather than on formatting, which is already settled by code.
Respond to Incidents: Carry on-call for the services you build, investigate and resolve incidents in your own systems, and turn root cause into a fix rather than a postmortem slide.
Publish What You Build: Put finished skills, agents, and components into DAT's capability marketplace so the next team that needs it finds it instead of rebuilding it.
The Skills and Experience You'll Bring
5+ years of experience in software engineering, with real production ownership and solutions you shipped, operated, and got paged for.
Demonstrated experience building with LLMs and agentic frameworks in production: tool-calling, retrieval-augmented generation, multi-agent orchestration (e.g., LangGraph, Bedrock Agents, MCP-style tool integration) which shipped and operated, not a weekend prototype.
Strong development skills in TypeScript/ js , Python, Java, or Go, with the discipline to write tests other engineers actually trust.
Solid, hands-on AWS experience: Bedrock,Agent Core Runtime, Strands, EKS/Kubernetes, Lambda, and S3 in production; working familiarity with Redpanda, MSK, or Kafka; comfortable in infrastructure-as-code (Pulumi, Terraform, or similar) even where you're not the one designing the account structure.
Practical experience building or maintaining evaluation and monitoring for AI systems: eval sets, drift and quality signals, and treating a model regression with the same seriousness as a code regression.
Comfortable working from a written specification rather than a verbal task. Ability to spot what's missing or ambiguous in a spec and push back before code, not after.
Experience authoring design docs, HLDs, or short technical papers that another engineer can execute against without you there to explain it.
Solid grounding in event-driven and distributed systems. You understand the tradeoffs Redpanda, MSK, kafka and relevant patterns force even if you're not the one drawing the topology.
Working knowledge of observability practices — dashboards, traces, logs, and alerts sufficient to diagnose your own service and argue for what it needs to hit our authored autonomy tiers.
Clear written and verbal communication — you'll be e

> History, not a promise: past H-1B filings do not mean an employer will sponsor you or a given role. Data: U.S. DOL OFLC LCA disclosure (FY2024–FY2026 for role counts, certified LCAs) and the USCIS H-1B Employer Data Hub. Not legal advice. Methodology: https://visajobs.io/methodology
