AIsa (https://aisa.one) is the unified resource and payment layer for AI agents. We give agents seamless access to models, data and payments—so builders can focus on what their agents do, not the infrastructure behind them. Backed by top VCs, we’re an agile, agent-enabled team moving fast.
The Role
You'll design and build AIsa's autonomous agents and multi-agent systems. These agents don't just call one model, they reason about cost, latency, and capability to dynamically route across LLM providers, orchestrate payments on behalf of users, and compose API calls on the fly. You'll architect the intelligence layer that turns AIsa from infrastructure into an autonomous platform.
Responsibilities
- Design and build autonomous agents and agent swarms that operate on top of AIsa's resource layer
- Implement dynamic LLM routing: cost-aware, latency-aware, capability-aware model selection at runtime
- Build agent-native payment flows where agents autonomously manage budgets, authorize spend, and settle transactions
- Architect dynamic API composition so agents can discover, chain, and call external services without hardcoded integrations
- Optimize end-to-end agent performance across multi-step reasoning, tool use, and orchestration
- Push forward our integration with agent protocols (MCP, A2A, Skills)
Requirements
- Strong Python skills
- Hands-on experience building with agent frameworks (LangChain, CrewAI, MetaGPT, or similar)
- Deep understanding of LLM capabilities, prompt engineering, and multi-model orchestration
- Familiarity with agent protocols and standards (MCP, A2A, Skills)
- Comfortable reasoning about cost/performance tradeoffs in LLM workloads
- AI-native development workflow (Cursor, Claude Code, etc.)
- Self-directed, fast learner, thrives in ambiguity
Nice to Have
- Experience with multi-agent coordination or swarm architectures
- Background in payment systems or autonomous transaction flows
- Experience building developer tools or SDKs
- Open-source contributions in the AI/agent space
- Familiarity with Go, TypeScript, or frontend development
Compensation