At TriumphPay we are building the transportation payments network for the future. Our software touches a combined $37.1B in annualized freight volume, representing over 20% of the brokered freight market in the U.S. TriumphPay’s customers are using our products to solve real world problems. This is both exciting and also an incredible responsibility.

We are looking for experienced Staff ML engineers to join our team of 35+ engineers. If you join TriumphPay, you will work closely in a small, cross-functional team of 3-4 people focused on our AI/ML systems. Our teams operate with a high degree of autonomy, allowing you to take ownership of projects from ideation to deployment. You’ll collaborate closely with product managers and other stakeholders to understand customer pain points and deliver impactful solutions that support critical features. 

Our engineering team is fully remote and believes strongly in work-life balance.

A Day In The Life: 

There’s no defined template that teams at TriumphPay follow, allowing each team to build the day that lets them perform at their best.

Typically, a team has a morning standup allowing them to catchup on what happened yesterday, and ensure there’s a plan in place for the day ahead. You’ll work with our product group and members of the sales team to ensure we’re building the tools our customers need to succeed.

The Tech:

The AI/ML team works primarily in a mixture of Python and Clojure for ML experimentation, data processing, and deployments, with Ruby and other languages used for integrating models into customer-facing applications. Python and Ruby make up the majority of our work, with Clojure being third.

Occasionally, the AI/ML team handles integration work in Ruby or other languages directly when it enables faster delivery of value, though this work may also be handed off to feature development teams. We use PySpark for most of our data processing, AWS SageMaker Studio for model development and validation, and PyTorch + HuggingFace for deep learning work. Model inference runs on a mix of FastAPI and Clojure applications, depending on the model type.

Our ML systems process more than 1 million documents per day through hundreds of models requiring robust pipelines to handle noisy, unstructured data with high precision at scale.  You'll work on building, deploying, and integrating models that can generalize across diverse document formats and adapt to evolving customer needs.  Our models must operate within strict latency requirements to ensure seamless customer experiences, while maintaining high performance in extracting and classifying data from complex, unstructured documents.  We are constantly exploring new techniques in deep learning, transfer learning, and model optimization to improve the accuracy and efficiency of our systems.

We know that good engineers can pick up new tools and languages on the job and we don't expect candidates to be familiar with all of these technologies.  We love curious individuals who believe they can always improve, and we know that good developers are capable of picking up new languages and tools.

Engineers are provided a top of the line MacBook to do their work, and you’ll have access to all the necessary tooling to do the non-coding parts of your job (Zoom, Slack, etc.).

To succeed in this role, you should be:

• Curious. You aren't content with the status quo and know that we can always improve.

• Data Driven.  You seek evidence to support hypotheses and identify optimal solutions within problem constraints accounting for sources of error and uncertainty.

• Collaborative. You can work with others to improve a solution iteratively factoring in new information from outside perspectives.

• Empathetic. Your designs are influenced by a deep understanding of the customers' needs.

• A strong communicator. You will proactively communicate issues and trade-offs with team members to support alignment and fast decision making.

• Be an outstanding developer. Your peers should recognize you as one of the best and the brightest developers they have worked with.

We hope you bring into this role:

• 10+ years of software engineering experience; 4+ years with production ML at scale.

• Proven success designing and operating distributed ML systems.

• Strong experience with model deployment patterns and data pipelines.

• Demonstrated ability to influence technical decisions across teams.

• Deep understanding of reliability engineering and production support.

Success at Triumph Looks Like:

• Leading multi-team ML initiatives that deliver sustained business value.

• Establishing clear, repeatable ML engineering standards adopted across teams.

• Reducing friction from experimentation to production.