AI/ ML Engineer

Job Description

AI Engineer

\n

New York City, NY (in person, relocation offered)

\n

\n

About the Company

\n

Our client builds mission-critical trust and safety infrastructure for some of the most important

\n

digital platforms in the world. Bad actors have always abused the internet, and AI is

\n

accelerating fraud, abuse, and manipulation at a scale and speed no human team can handle

\n

alone. Their platform gives trust and safety teams a single place to write and enforce policy,

\n

deploy AI agents against abuse in real time, investigate threats, handle regulatory reporting,

\n

and measure whether their safety programs are working.

\n

\n

Customers include some of the largest internet platforms in the world, and decisions made in

\n

the software directly determine what stays up, what comes down, and how users are treated.

\n

\n

The company is a small, fast-moving, venture-backed team that values being intentional,

\n

direct, and deeply focused on solving real customer problems.

\n

\n

Why This Role

\n

The company is growing quickly and is adding an AI Engineer to partner with the existing AI

\n

Engineers, Data Scientist, and Data Engineer. You’ll build the ML systems that make the

\n

platform faster, more efficient, and more accurate, turning the enormous volume of customer

\n

decision data it processes into production models that directly shape customer outcomes.

\n

\n

We’re looking for a builder. Someone who has taken models from messy data to production at

\n

scale, who reaches for a gradient-boosted tree before a transformer when that’s the right call,

\n

and who has stood up ML infrastructure from scratch rather than inheriting it fully built. What

\n

matters most is judgment: knowing the smallest, most efficient, most reliable model for the

\n

job, and understanding both ML methods and LLMs deeply enough to make that call yourself

\n

rather than defaulting to whichever one you know best. This role needs someone who cares

\n

as much about precision-recall tradeoffs, class imbalance, and serving latency as they do

\n

about model architecture.

\n

\n

What you’ll do

\n

\n
• Turn real-world customer data into something a model can actually learn from, then
\n

\n

decide what model approach fits: a classical classifier when it wins on cost and latency,

\n

a fine-tuned LLM when the tradeoff is worth it, a third-party API as a bootstrap. You own

\n

the full path from data to decision, not just the model.

\n

\n
• Improve the classification pipeline, confidence cascading, and detection strategies so
\n

\n

harmful content is caught efficiently, balancing cost, latency, and accuracy deliberately.

\n

\n
• Develop intelligent features that help moderators make decisions, organize platform
\n

\n

content, and reveal patterns across the data.

\n

\n
• Partner with Engineering to build out the in-house model training, hosting, and
\n

\n

inference platform.

\n

\n
• Design and build the evaluation and metrics infrastructure customers rely on, including
\n

\n

how classifier scores and model outputs are calculated, stored, surfaced, and iterated

\n

on.

\n

\n
• Partner with the Founding Data Scientist and AI Engineers to shape the agent
\n

\n

evaluation architecture, measuring whether the agent fleet is making the right decisions

\n

with the right tools at the right cost.

\n

\n
• Partner with the Data Engineer to shape the data infrastructure powering the ML
\n

\n

systems, ensuring model training, feature pipelines, and production inference have the

\n

right data flowing at the right latency and scale.

\n

\n
• Mentor teammates and raise the ML bar across the company as its ML capability
\n

\n

matures.

\n

\n

What we’re looking for

\n

\n
• 5 to 8+ years of machine learning engineering experience on a small team, with a
\n

\n

strong track record of shipping ML systems (gradient boosting, tree-based models,

\n

classifiers, embedding-based methods) to production.

\n

\n
• You’ve taken a classification problem from messy, unlabeled, real-world data all the
\n

\n

way to a model that shipped and served production traffic.

\n

\n
• You understand LLMs well enough to make an informed, defensible call about when an
\n

\n

LLM is worth its cost and latency versus a classic model.

\n

\n
• Real, hands-on experience building classifiers under severe class imbalance, where
\n

\n

the signal you care about is a small minority of the data.

\n

\n
• Thrived in environments where the ML infrastructure wasn’t already built for you: you’ve
\n

\n

stood up training pipelines, serving infrastructure, evaluation harnesses, and monitoring

\n

from scratch rather than inheriting a mature platform.

\n

\n
• Startup or small/mid-size company experience where you owned meaningful scope and
\n

\n

had to make pragmatic tradeoffs about what to build, what to buy, and what to defer.

\n

\n
• Deep fluency with the fundamentals: thoughtful feature engineering, leak-aware
\n

\n

train/test splits, metric selection on imbalanced data (precision/recall/F1/AUC over

\n

accuracy), cross-validation, and principled hyperparameter tuning.

\n

\n
• Strong Python skills and hands-on experience with AI and ML frameworks (PyTorch,
\n

\n

scikit-learn, LangChain, XGBoost, etc.).

\n

\n
• Solid MLOps foundation: CI/CD for ML, model versioning, experiment tracking, drift
\n

\n

detection, and production monitoring. Bonus if you have experience with training,

\n

evaluating and serving models via Databricks.

\n

\n
• Experience designing inference systems with explicit latency and throughput targets,
\n

\n

and independently making informed tradeoffs between model complexity, cost, and

\n

performance.

\n

\n
• Experience with AWS and infrastructure-as-code (Terraform) is a plus.
\n

\n

\n

Location & Benefits

\n

This role is based in New York City, and relocation is offered. The team works together in

\n

person and holds at least two all-company events per year. Benefits include health, vision and

\n

dental coverage, a 401(k) plan with employer matching, fully paid commuter benefits, and a

\n

fully stocked office with paid lunch and dinner.