ML Data Platform Engineer

Parisi Labs•New York, NY

2d•$175,000 - $245,000•Hybrid

About The Position

Parisi Labs is an AI company building learning systems for complex physical environments. We combine historical and live data with real operational context to help people understand the present, evaluate possible futures, and make better decisions. Energy is our first proving ground. Ask The Grid (https://askthegrid.com) is our public product for exploring the systems, markets, and assets that make up the power grid. We are a small technical team working across machine learning, data infrastructure, software, and real-world operations. We are looking for an ML Data Platform Engineer to make the data behind our models and products dependable, understandable, and easy to use. This role sits where data engineering meets machine learning. You will turn messy, changing real-world sources into durable datasets and interfaces that researchers, product engineers, and customer-facing technical teams can trust. The goal is not to build a large platform for its own sake. It is to make each new model, product surface, and authorized data source faster to bring online without compromising correctness.

Requirements

Enjoy making difficult real-world data useful, not merely moving it between systems.
Understand how warehouse or streaming data becomes training, evaluation, and product data.
Care about temporal correctness, reproducibility, leakage, lineage, and source rights.
Ability to design practical systems without reaching immediately for a large-company platform.
Comfortable debugging incomplete APIs, changing schemas, and surprising data behavior.
Communicate clearly with researchers, product engineers, and customer-facing teammates.
Desire substantial ownership on a small team and ability to make progress without a mature data organization around you.
Strong Python and SQL experience.
Experience with data engineering, ML data systems, dataset engineering, platform engineering, or high-quality analytics engineering.
Experience with object storage, warehouses, streaming or workflow systems, columnar formats, APIs, and data-quality tooling.
Experience preparing data for model training, evaluation, or scientific computing.

Nice To Haves

Familiarity with time-series, geospatial, weather, market, event, or other temporally sensitive data is useful.
Startup or small-team experience is helpful, but evidence of unusually strong ownership matters more than a particular company background.

Responsibilities

Build and improve ingestion, backfill, validation, and observability for high-volume, time-dependent data.
Define clear data contracts and point-in-time semantics for model training, evaluation, and product use.
Create reusable workflows for bringing new public and customer-authorized sources into the system.
Build quality, lineage, freshness, and access controls that make data trustworthy in repeated use.
Develop efficient datasets and query interfaces for machine-learning and product workloads.
Diagnose whether failures originate in source data, transformations, model inputs, or serving systems.
Work closely with research and product engineers so data requirements become reliable software, not recurring manual projects.
Exercise judgment about which abstractions should become durable infrastructure and which should remain purpose-built.