Delivery Engineer, ML, Americas

Hilbert; Lima Metropolitana; Remote; PEN 416,000 - 624,000; Full time

Job summary

Hilbert is a data science‑first growth engine helping B2C teams understand user behavior and revenue drivers. You will build models, deploy them in customer environments or our stack, and measure uplift across time zones while collaborating with global teams.

You’ll own feature development, model evaluation, and production monitoring, partnering with product and data delivery engineers to ship repeatable solutions for varied customers.

Qualifications

• Extensive applied ML experience on tabular customer data.
• Deep learning experience designing architectures and training models with PyTorch or similar.
• Feature engineering in SQL warehouses such as Databricks, Snowflake or BigQuery.
• Production ML with tools such as MLflow, Weights & Biases or Vertex AI.
• Strong Python and ML libraries.
• Experience working in distributed teams across time zones.
• Fluent Spanish for roles based in Latin America.

Responsibilities

• Feature stores: build customer‑specific, at‑time correct features on the data foundation.
• Models: train, tune and validate propensity, churn, LTV, uplift and custom models.
• Model analysis: verify what each model learned and how it behaves across segments.
• Deployment: integrate models into customer environments or managed stack.
• Measurement: own power analysis, holdout setup, uplift analysis.
• Trust: document models for data science and risk teams; monitor in production.
• Productization: turn recurring features into product with engineering team.

Skills

Applied ML (tabular data)

Python

Deep learning (PyTorch)

Feature engineering

SQL/Data warehouses

Model deployment

Experiment design

Distributed teams

Spanish language

Tools

Databricks

Snowflake

BigQuery

MLflow

Weights & Biases

Vertex AI

PyTorch

Job description

About Hilbert

Hilbert is a scalable, data science-first growth engine that gives B2C teams predictive clarity into user behavior, revenue drivers, and the actions that drive sustainable growth. Fully agentic by design, Hilbert shrinks months-long decision cycles to minutes.

About Hilbert

Hilbert is a scalable, data science-first growth engine that gives B2C teams predictive clarity into user behavior, revenue drivers, and the actions that drive sustainable growth. Fully agentic by design, Hilbert shrinks months-long decision cycles to minutes. From Fortune 10 enterprises to beloved brands like FreshDirect, Blank Street, and Levain Bakery, operators run their growth on Hilbert. We're also co-building alongside leading AI companies.

The Mission: Turn Ground Truth Into Growth

Hilbert has two parts: a data foundation that gives our AI agents a high-fidelity view of each business, and an agentic web app where growth teams find and act on insights. The scores and audiences you build feed both: our agents reason with them, and our activation sends them into paid media and CRM.

The hard problem

Every customer's data is different, their compute may be ClickHouse, Databricks, Snowflake or BigQuery, and their teams want proof, not promises. Your models have to work in that environment and show measured uplift. We want every pod to stay lean and efficient across very different customer environments, delivering quality work at startup speed. You focus on delivery; our sales engineers handle pre-sales. Your work centers on customers in your own region: supporting customers elsewhere is a rare exception, and your global work is mainly staying aligned with Hilbert colleagues in other regions.

What You'll Do

• Feature stores: build customer‑specific, point‑in‑time correct features on the data foundation.
• Models: train, tune and validate propensity, churn, lifetime value, uplift and custom models, starting from our ML templates and building custom models when a customer needs something the templates don't cover. New model types that many customers will need are built first by our product engineering team, and you roll them out to later customers.
• Model analysis: go beyond aggregate metrics to verify what each model has learned, what drives its scores, and how it behaves across segments and on edge cases.
• Deployment: models run in the customer's own environment or on our managed stack. In either case, understand the constraints, pin down the requirements, design the ML system to fit, and write the scores and audiences our agents and activation use.
• Measurement: own the statistical design inside our growth team's measurement plan, from power and holdout setup to uplift analysis.
• Trust: document models for the customer's data science and risk teams, then monitor and retrain them in production.
• Productization: turn repeated features and model patterns into product with our product engineering team.

What Success Looks Like

• Audiences built from your models lift the agreed growth metric, such as conversion or retention, against a holdout group, and that result decides whether the customer takes them to production.

Who You Are

• The translator: you explain a model to a customer's data science team and earn their trust.
• The experimenter: you'd rather prove uplift in a holdout test than show off an offline metric.
• The pilot: you get the most out of AI agents across model iterations, from feature ideas and experiment runs to error analysis, and help build them into how the team works.
• The systems thinker: you build reusable features, pipelines and automation instead of one‑off models, so a small team can deliver across very different environments.
• Extensive applied ML experience on tabular customer data, with strong Python and ML libraries.
• Extensive hands‑on deep learning experience, from designing architectures and loss functions to training and shipping models in production, with PyTorch or a similar framework.
• Feature engineering in SQL warehouses such as Databricks, Snowflake or BigQuery.
• Working data engineering experience: building and debugging pipelines, data orchestration, data modeling and data quality checks, enough to own your feature pipelines end to end with our data engineers.
• Production ML with tools such as MLflow, Weights & Biases or Vertex AI.
• Solid experiment design and uplift measurement.
• Comfortable working in a distributed team across time zones.
• Fluent Spanish for roles based in Latin America.

Nice to have

• DevOps, software architecture and security knowledge
• Growth‑focused data projects experience
• Industry experience such as e‑commerce, retail, fintech, telco and media
• Model governance in regulated industries such as banking and telco: model documentation, explainability and PII handling
• Able to travel to the US when necessary

Reporting and location

You report to your pod's Lead Delivery Engineer and work alongside our data delivery engineers and product engineering team. The role is based in Americas. Delivery runs remotely, with occasional visits to our largest customers.

Why join us

At Hilbert, we move fast, work collaboratively, and give people real ownership over their impact. As a fast‑growing company, there's no shortage of room to grow — you'll take on new challenges quickly and shape the path as you go.

Hilbert is an Equal Opportunity Employer. We are committed to building a diverse team and an inclusive culture, and we welcome applicants of all backgrounds. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, gender identity or expression, sexual orientation, national origin, age, disability, veteran status, or any other status protected by applicable law.