Full Stack Developer

Data centers run on thousands of sensors, meters, and control systems, and the teams who run their power infrastructure have to make sense of all of it in real time. Teserac is building Neuron™, an AI-native observability and automation platform that gives those teams full visibility into their power operations and helps them get ahead of problems. At its core is an AI teammate that works alongside every operator around the clock: detecting anomalies, correlating events, and recommending what to do next.

As a Full Stack Developer, you'll own features end to end, from the ingestion path that pulls telemetry off a BMS in a live facility, through the time-series layer that stores it, to the dashboard an operator is watching at 2am.

This is a small team with real customers in production. Your code ships in days, not quarters, and it runs in facilities where being wrong has physical consequences. You'll work directly with the Principal Software & Architecture Lead, Product, and Solution Engineering. We move fast, avoid bureaucracy, and expect engineers to own outcomes rather than tickets.

This is a hybrid role based in our Santa Clara office, and visa sponsorship is available.

You're a full-stack engineer who is comfortable on both sides of the line, and who gets interested rather than annoyed when the data is messy, the vendor's API is undocumented, and the protocol spec is a PDF from 2009. You likely:

• Build front to back. You can trace a value from a Modbus register to a rendered chart without handing off, and you know where it's likely to break.
• Take databases seriously. You know that a query plan, an index, and a migration on a large table are engineering decisions, not configuration.
• Have worked with real time-series data at a volume where naive approaches stop working, and understand what that costs at read time and write time.
• Debug production calmly. When CPU is pinned and a queue is backing up, you reason from evidence rather than restarting things.
• Use AI tooling well. You get real leverage from Cursor, Claude Code, or similar, and your standards didn't drop when you started using it.

• Build operator-facing features for Neuron™ in React and Type Script: real-time dashboards, alerting, alarm and ticket workflows, capacity planning views, and power-flow visualizations.
• Build the interfaces the AI teammate works through: interaction panels, recommendation displays, approval workflows, and audit trails.
• Keep large, data-dense views fast and responsive across devices.

• Build and maintain the ingestion path that brings telemetry from customer sites into Neuron™: edge collectors, MQTT over AWS IoT Core, queue-based workers, and normalization into a canonical metric model.
• Write integration modules against real building and power systems:
  Siemens Desigo, Schneider PME and Eco Struxure, Niagara, Eaton, and generic BACnet, Modbus, and SNMP sources.
• Own the correctness of what lands: point mapping, unit and phase semantics, deduplication, and backfill.

• Develop against PostgreSQL and Timescale DB: hypertables, continuous aggregates, compression, and retention.
• Design schema and migrations that stay safe as tables grow into the billions of rows.
• Find and fix the performance problems that only appear with real customer data volumes.

• Write tested, reviewed code, and contribute to CI/CD and automated testing.
• Troubleshoot production issues across the stack, and share on-call as the platform scales.

We're not building incremental improvements to legacy tools. We're re-architecting how data centers see, understand, and act on their own operations.

• Ship features that…