Software Engineer, Data Platform Lead (Horizon AI)
About The Position
Horizon AI is building an operating system for how organizations run themselves, beginning with autonomous data centers. A fleet of AI agents monitors STN’s GPU infrastructure, reasons about its state, and increasingly takes action. The system is already running in production on real fleet data. Actions with bounded risk and verifiable outcomes are handled by agents, while riskier cases are presented to humans as recommendations. The work is to move decisions into autonomous action incrementally, based on evidence. Data centers are the proving ground for capabilities intended to generalize beyond infrastructure. The Data Platform Lead will build the data foundation used by the platform’s agents. Fleet information is currently distributed across a relational store, a DCIM system, and multiple time-series and monitoring systems, each with a separate query path and no unified model. There is also no systematic pipeline for capturing what operators observed, concluded, and chose, although that record is needed to support broader agent autonomy. Retrieval is currently document-oriented rather than fleet-state-oriented, and telemetry retention and granularity are tuned for dashboards rather than historical learning. The role will create a queryable fleet knowledge model covering assets, topology, configuration, and current state; build a decision-capture pipeline and instrument operator workflows to collect labelled human decisions; and own the data-access layer that abstracts telemetry sources. The lead will negotiate telemetry requirements with source-owning teams, including signals, granularity, retention, and raw-event access, and own the scale path, including vector-index behavior and eventual separation of analytics as the schema grows.
Requirements
- Strong data-modelling judgment, particularly for making physical or infrastructure domains queryable.
- Experience with time-series and event data at scale, including cardinality and retention trade-offs.
- Depth in PostgreSQL, including vector search, and comfort working with a large, evolving schema.
- Sufficient infrastructure literacy to understand fleets, including nodes, GPUs, fabric, schedulers, and power.
- Bias toward shipping a usable model early rather than designing a complete model first.
Nice To Haves
- Familiarity with DCIM tooling.
- GPU fleet or HPC exposure.
- Experience with DCGM.
- Experience with Slurm.
- Experience with entity resolution across inconsistent sources.
Responsibilities
- Design and build a fleet knowledge model covering assets, topology, configuration, and current state, queryable by an agent through one call.
- Build a decision-capture pipeline and instrument operator workflows to accumulate labelled human decisions.
- Own the data-access layer that abstracts telemetry sources, enabling continued use of current sources and the option to build an internal collection if needed.
- Negotiate telemetry requirements with the teams that own the sources, including signals, granularity, retention, and raw-event access.
- Own the scale path, including vector-index behavior and the analytics separation that a growing schema will eventually need.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Job Type
Full-time
Career Level
Principal
Education Level
No Education Listed
