Senior/Staff Founding Full-Stack Engineer, Agent Systems
Cephas Consultancy Services Private Limited
United States, California, San Francisco
Headcount: 1
Job Details
- Experience: 4+ years
- Salary: Not specified
- Equity: 1%–2%
- Visa sponsorship: H-1B, O-1, OPT
- Employment type: Full-time
- Work type: In person
Benefits & Perks
- Meals in the office
- Health insurance
- Unlimited PTO
About This Role
This is a backend- and systems-heavy founding engineering role where “full-stack” means owning the product and system end to end. The role is based at our San Francisco office, in person five days a week.
What You’ll Own
- The workflow engine for long-running human and agent work: durable state, retries, idempotency, approvals, replay, compensation, and auditability
- Identity and authorization for humans, services, and agents: principals, delegation, scoped sessions, tenant isolation, and traceable authority
- A versioned policy engine defining who or what can perform which action on which resource, and under what context
- Secure storage and use of customer keys, OAuth tokens, and admin credentials, including isolation, rotation, ephemeral injection, revocation, and blast-radius containment
- The probabilistic-to-deterministic execution boundary, including validation, evaluations, and release gates
- Isolated virtual environments and development boxes for agents, including reproducible state, observability, resource controls, and safe teardown
- The interfaces and APIs to author, approve, inspect, debug, and operate these systems in production
Requirements
- Experience designing and operating complex production systems, with the ability to explain what broke, how it was recovered, and what changed
- Strong backend and systems architecture skills, with the ability to work across frontend, data, infrastructure, and product
- Depth in one or more of the following: IAM/authorization, policy systems, workflow orchestration, security, multi-tenant SaaS, agent infrastructure, MDM, ITSM, enterprise integrations, or virtualized execution
- Thinks in invariants, threat models, failure modes, and user outcomes
- Works well amid ambiguity and moves quickly without compromising correctness, security, or operability