Backend Engineer, Models

Networking is one of the most fundamental industries in all of technology. For the first time, Meter has unified the full networking stack; and now we are making it autonomous.

We are building a neural network-driven system for reasoning in raw computer networks to solve any and all networking problems. As described on Meter.ai, we’re building models in a closed-loop system that takes (as input) real-time telemetry, logs, and events on the network to autonomously troubleshoot, improve performance, and resolve issues.

To make this possible, we don’t just need great models; we need infrastructure that gives those models clean, versioned, low-latency access to the right data, across training, evaluation, and deployment.

Every Meter network deployed in the field is a rich data source for our Models team. But without careful infrastructure design, this data becomes fragmented, stale, or inconsistent. Your job is to make sure that never happens. You’ll own the core data interface that powers our model development, experimentation, evaluation, and real-time inference.

This is a foundational role with outsized impact. Your work will define how quickly we can train new models, how reliably we can evaluate them, and how seamlessly they can operate in production across hundreds of real-world networks. You’ll partner tightly with modelers to ship systems that feel elegant, scalable, and bulletproof.

• →

  Design and implement the Models API: a unified interface for accessing training, evaluation, and deployment data across raw, transformed, and feature-engineered layers

• →

  Ensure backwards compatibility and feature versioning across constantly evolving schemas

• →

  Build scalable pipelines for ingesting, transforming, and serving petabytes of data across Kafka, Postgres, and Clickhouse

• →

  Create CI/CD workflows that evolve the API in lockstep with changes to the underlying data schema

• →

  Enable fine-grained querying of historical and real-time data for any network, at any point in time

• →

  Help define and enforce the principle of "smart data, dumb functions": doing as much as possible in the data layer to keep downstream code minimal

• →

  Collaborate with modelers to co-design training and evaluation pipelines that are reproducible, debuggable, and fast

• →

  Own performance across key endpoints to meet real-time serving constraints

• Have experience designing large-scale data infrastructure, ideally across batch and streaming modes

• Think deeply about schema design, versioning, and data quality

• Care about making systems simple to use and hard to misuse

• Enjoy partnering closely with research or modeling teams

• Thrive in early-stage environments and like building from scratch

• Kafka, Postgres, Clickhouse

• Python, Go

• AWS, Azure