Instabase
Software Engineer - Agent Harness
About this role
Instabase is seeking a Staff AI Engineer to design and architect the Agent Harness, a secure runtime execution engine that powers their SuperApp platform. You'll build the state machines, sandboxed execution environments, and tool-calling frameworks that enable AI agents to safely execute code and interact with external systems at enterprise scale.
What you'll do
- Architect and build the Agent Harness execution runtime for planning, reasoning, memory, and tool-execution loops with reliable state management
- Design secure, isolated sandboxed environments (Docker, gVisor, WebAssembly, microVMs) to safely execute agent-generated code
- Create high-throughput API and integration layers connecting SuperApp to external services, databases, and tools
- Engineer system-level security guardrails to prevent prompt injection, jailbreak attempts, and enforce data isolation
- Optimize LLM token usage and reduce latencies through context caching, routing, and window pruning strategies
- Lead technical strategy through design documents and mentor senior/mid-level engineers across the organization
What they're looking for
- Distributed systems design and operation
- Agent execution environments and LLM orchestration
- Runtime isolation and sandboxing technologies
- Backend systems architecture
- Security and guardrail engineering
- Tool-calling frameworks and integration patterns
- Context optimization and caching strategies
- Technical leadership and mentorship
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Instabase
Instabase builds an AI-powered platform for automating the processing of unstructured data into actionable insights through intelligent workflows and no-code automation. The company is hiring Solutions Engineers and Sales Engineers to design, deploy, and architect custom AI solutions for enterprise customers.
- Website
- instabase.com
Likely interview questions
- Walk us through your experience designing and operating distributed systems at scale—what was your largest production challenge and how did you solve it?
- Tell us about a time you built or significantly modified an agent execution environment or LLM orchestration system. What were the key architectural decisions?