Research Scientist

Auger•Bellevue, WA

22d

About The Position

Build at Auger Auger is the autonomous operating system for supply chains — the layer that finally allows disparate systems like ERP, WMS, and TMS to work together instead of against each other. Most supply chain software surfaces problems and waits for a human to act. Auger solves them. Our AI detects disruptions, evaluates trade-offs, and executes decisions automatically — moving from signal to action in seconds, not weeks. We eliminate the Coordination Tax: the billions in capital and time lost when disconnected systems force the best people in the business to become the Human API between planning and execution. At Auger, we design autonomy into our systems. We expect the same from our people. That means: Clear ownership, not decision by consensus First principles over inherited patterns Shipping systems, not slide decks Fast feedback from reality, not opinions If you want to build, ship, and iterate against reality, Auger is for you. Auger was founded by Dave Clark and is backed by $150M from Oak HC/FT and Eclipse Capital. Headquarters in Dallas, TX and Bellevue, Washington. About the Team & Role The core of this work is training foundational models for supply chain expertise, not wrapping a generalist frontier model in a better prompt. We think a specialist model, trained deep on the domain, beats a generalist model on the problems that actually matter here, and gets us to a level of inference speed, cost, and reliability that routing every decision through a frontier API simply can't reach.

Requirements

Built foundational training data at scale, corpora and not just models.
Understand that what goes into a model matters as much as its architecture.
Led a project across multiple release cycles, each one measurably better than the last.
Designed evaluation methodology that goes beyond standard benchmarks, built specifically to surface what those benchmarks miss.
Adapted general purpose models to specialized, knowledge intensive domains and understand what actually transfers versus what has to be rebuilt.
Created datasets that other researchers and practitioners now build on.
Taken research past the paper and into a real, end to end system that people other than researchers actually use.
Recognition, best paper or outstanding paper or otherwise, has followed the work, but wasn't the point of the work.

Nice To Haves

The instinct to Explore to Evolve.
The ability to treat a bad eval run as signal, not shame, and have a short loop from identifying issues to the next checkpoint.
The ability to Crush Complexity by shipping clean datasets and clean evals that teammates can easily use.
Experience with tech-leading through v1, v2, v3, each release measurably stronger than the last.
Understanding that the job is never actually done, there's always a v4.

Responsibilities

Training foundational models for supply chain expertise.
Deriving why techniques work and rebuilding pipelines from the tokenizer up when the domain demands it.
Developing evaluation frameworks to catch what standard benchmarks miss.
Shipping clean datasets and clean evaluations.
Iterating on releases, with each release measurably stronger than the last.
Watching checkpoints run flawlessly in production, under real load, on real customer data.