ego

18 hours ago

Member of Technical Staff: Agent Runtime

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI

Last checked: 11 hours agoStatus: Live

Job Summary

Own the core agent runtime that powers Ego’s AI companions and agents. Build a reliable agentic loop as a state machine, handling tool-calling errors safely and ensuring durable, checkpointed sessions with idempotency and clear execution boundaries. Lead long-horizon context compaction, design an extensible tool registry, and ship deployable systems with clean APIs, persistence (Postgres/SQLite/Cloudflare), and observability for per-user durable agent instances.

Location: San Francisco, Tokyo, Singapore

Workplace: Hybrid

Employment Type: Full time

Job Function: Software Engineering

Seniority: Mid level

Key Responsibilities

•Design and maintain the core agentic loop (plan → tool call → observe → repeat → finalize) as a resilient state machine.
•Build durable, checkpointed, resumable sessions with at-least-once vs exactly-once semantics and crash safety.
•Own long-horizon context compaction (pinned goals, summaries, sliding windows, and retrieval over older state).
•Create an extensible tool registry with JSON schema validation and swappable model adapters.
•Ship deployable runtime components, including local bring-up, persistence interface (Postgres/SQLite/Cloudflare), clean HTTP APIs, and per-user durable agent instances.

Pay and Benefits

Salary: USD 100,000 - 160,000 annually

Equity and Bonus:Equity

Key Requirements

•3+ years building backend or infrastructure systems with real production ownership of stateful services.
•Hands-on experience building LLM agent systems, including tool-calling loops, defensive handling of model output, and step/token/cost budgets.
•Deep grasp of durable execution: checkpointing, idempotency, outbox/step-log patterns, and reasoning about failure windows.
•A deliberate context-management strategy for long-horizon tasks, with the ability to defend trade-offs.
•Strong API and data-model design; persistence behind clean seams; comfort being judged on docker compose up working from a README.

Experience:3+ yearsLLM agentsBackend infrastructure

Skills:Written communication

Tech Stack:LLMPostgresSQLiteCloudflare Durable ObjectsCloudflare WorkersD1Docker ComposeHTTPJSONWebhooksCronIdempotency keysStep logsObservabilityTracing

Company Brief

ego

Develops personalized AI assistants and agent tooling that connect to users' apps and data to automate tasks, surface insights, and augment workflows. Offers APIs and SDKs for integrating proactive, context-aware AI capabilities into products and services.

Industry: AI & Machine Learning

Company Size: Small (11 to 50 employees)

Growth: Early Stage Startup

Funding: Seed

Headquarters: San Francisco, United States

Website

Loading job details...

Preparing the role view and application actions.

ego