AI/LLM Engineer
Overview
We’re looking for an adaptable AI/LLM engineer who thrives in ambiguity—someone comfortable jumping into new problem spaces, rapidly learning what’s needed, and shipping pragmatic solutions. You’ll turn fuzzy ideas into working prototypes and then into reliable, scalable systems that create real value.
What you'll do:
Tackle open-ended AI problems: clarify goals, propose approach options, and choose sensible trade-offs.
Stand up end-to-end workflows—from data wrangling and evaluation through deployment and monitoring.
Build quick experiments and MVPs to de-risk unknowns, then harden them for production.
Create lightweight tooling that helps others explore, test, and iterate on AI features.
Work across teams (security, infra, product, domain experts) to ship responsibly in real-world environments, including sensitive contexts.
Document decisions, assumptions, and risks so others can build on your work.
What you bring:
U.S. citizenship.
Solid software fundamentals and strong Python skills; you write clear, maintainable code and tests.
2–5 years of hands-on experience building and shipping ML/AI or NLP-driven features (titles less important than impact).
A generalist mindset: you can learn unfamiliar libraries, models, or stacks quickly and pick the right level of sophistication for the problem.
Practical evaluation chops: you design metrics, create test sets, and know when something is “good enough” to pilot vs. needs more rigor.
Data instincts: you’re comfortable sourcing, cleaning, labeling, and shaping both structured and unstructured data.
Bias for action and ownership in fast-moving, resource-constrained settings.
Thoughtful approach to safety, privacy, and policy constraints.
Nice to have:
Experience building knowledge- or retrieval-oriented applications.
Exposure to edge or low-resource deployments.
Comfort interfacing with stakeholders and non-technical users.
How we work:
Start small, learn fast: prototype, measure, iterate.
Prefer simple, observable systems over “state-of-the-art” complexity.
Write things down; leave a trail others can follow.
AI/LLM Engineer
Overview
We’re looking for an adaptable AI/LLM engineer who thrives in ambiguity—someone comfortable jumping into new problem spaces, rapidly learning what’s needed, and shipping pragmatic solutions. You’ll turn fuzzy ideas into working prototypes and then into reliable, scalable systems that create real value.
What you'll do:
Tackle open-ended AI problems: clarify goals, propose approach options, and choose sensible trade-offs.
Stand up end-to-end workflows—from data wrangling and evaluation through deployment and monitoring.
Build quick experiments and MVPs to de-risk unknowns, then harden them for production.
Create lightweight tooling that helps others explore, test, and iterate on AI features.
Work across teams (security, infra, product, domain experts) to ship responsibly in real-world environments, including sensitive contexts.
Document decisions, assumptions, and risks so others can build on your work.
What you bring:
U.S. citizenship.
Solid software fundamentals and strong Python skills; you write clear, maintainable code and tests.
2–5 years of hands-on experience building and shipping ML/AI or NLP-driven features (titles less important than impact).
A generalist mindset: you can learn unfamiliar libraries, models, or stacks quickly and pick the right level of sophistication for the problem.
Practical evaluation chops: you design metrics, create test sets, and know when something is “good enough” to pilot vs. needs more rigor.
Data instincts: you’re comfortable sourcing, cleaning, labeling, and shaping both structured and unstructured data.
Bias for action and ownership in fast-moving, resource-constrained settings.
Thoughtful approach to safety, privacy, and policy constraints.
Nice to have:
Experience building knowledge- or retrieval-oriented applications.
Exposure to edge or low-resource deployments.
Comfort interfacing with stakeholders and non-technical users.
How we work:
Start small, learn fast: prototype, measure, iterate.
Prefer simple, observable systems over “state-of-the-art” complexity.
Write things down; leave a trail others can follow.