Founding Agentic AI Engineer

Job summary

Design and build Earthinia’s agentic AI runtime, including task routing, multi-agent orchestration, tool execution, memory, retrieval, integrations, permissions, and human approval workflows. Establish reliable evaluation, recovery, tracing, security, and performance practices while shaping the platform’s technical direction with the founder.

Job details

Earthinia is building an enterprise AI workforce platform that enables businesses to create AI-powered operational teams composed of specialized roles, tools, knowledge, permissions, and integrations. We are looking for a Founding Agentic AI Engineer to join at an early stage and take significant technical ownership of Earthinia's agentic intelligence and execution infrastructure. This is not a conventional AI engineering position. You will work directly with the founder to design and build the systems that allow Earthinia clusters and instances to understand requests, select the appropriate capabilities, use tools safely, collaborate across roles, maintain context, and execute real business workflows. What You Will Own Design and develop Earthinia's agentic AI runtime. Build intelligent task routing between clusters and specialized instances. Develop multi-agent planning, orchestration, delegation, and execution systems. Build reliable LLM tool-calling and function-execution infrastructure. Design agent-to-agent communication and multi-step workflows. Implement model abstraction and routing across supported AI providers. Develop context management, conversation memory, long-term memory, and knowledge retrieval. Build RAG pipelines for customer and instance-specific knowledge. Design MCP-based and native integrations with external systems. Implement permission-aware tool execution and security boundaries. Develop human-in-the-loop approval workflows for sensitive actions. Build retry, fallback, failure-recovery, and execution-state mechanisms. Create evaluation frameworks for measuring agent accuracy, reliability, tool selection, and task completion. Build tracing and observability for debugging complex agent executions. Optimize model usage, latency, token consumption, and inference cost. Help establish engineering standards for testing, reliability, security, documentation, and maintainability. Work closely with the founder and future full-stack engineers to expose agentic capabilities through the Earthinia Console, API, SDKs, and communication channels. Example of What You Will Build A business should eventually be able to send a request such as: "Analyze our strongest sales opportunities, prepare personalized outreach, schedule follow-ups, and update our CRM." Earthinia should determine which specialized instances are required, identify their permitted tools, coordinate execution between them, preserve context, request approval where necessary, execute the workflow, and return a traceable result. You will help build the intelligence that makes this possible. What We're Looking For We value demonstrated engineering ability more than credentials. You should be comfortable building systems from first principles in an early-stage startup where architecture, requirements, and product decisions evolve quickly. Strong experience in several of the following areas is preferred: Python FastAPI Agentic AI systems LLM APIs Tool/function calling Multi-agent systems MCP Retrieval-Augmented Generation Vector search and embeddings Prompt and context engineering Agent memory architectures OpenAI, Anthropic, or other model APIs PostgreSQL Redis AWS Docker Kubernetes Distributed systems Event-driven architectures API design AI evaluations and observability Experience with frameworks such as LangGraph, LangChain, LlamaIndex, Semantic Kernel, or similar systems is useful, but we care more about your ability to understand and engineer the underlying architecture than dependence on any particular framework. You May Be a Strong Fit If You Have built production AI agents rather than only chatbot prototypes. Understand the difference between an LLM call and a reliable agentic system. Can reason about tool selection, permissions, memory, planning, orchestration, and failure recovery. Have integrated AI systems with external APIs or business applications. Can design clean, reusable, maintainable Python systems. Are comfortable making architectural decisions with incomplete information. Think carefully about security and reliability when AI systems can take actions. Can work independently with significant ownership. Want to help build a company and platform rather than simply complete assigned tickets. Early-Stage Environment Earthinia is at an early stage. This role provides substantial ownership and influence over the technical direction of the company. You will work closely with the founder and become one of the earliest members of the engineering organization. This is a hands-on position. We are looking for a builder who wants to write code, experiment, solve difficult agentic AI problems, speak directly about architectural tradeoffs, and help turn Earthinia into a production-grade platform. Compensation This position is currently equity-only. There is no cash salary at the current stage. We want candidates to understand this before applying. The expected equity range is: 2%–5% ownership, depending on experience, technical depth, commitment, responsibility, and expected long-term contribution. Equity is expected to follow a standard: 4-year vesting schedule 1-year cliff Monthly vesting thereafter The final grant will be formally documented through the appropriate company equity agreements once legally available and will be subject to applicable corporate approvals and company formation/equity documentation. As Earthinia raises financing or develops sufficient recurring revenue, the company intends to introduce competitive cash compensation. Future financing or salary, however, should not be treated as guaranteed compensation when accepting this position. This opportunity is best suited to someone who understands the risk of joining an early-stage company and wants meaningful ownership in the platform they help create.

What you’ll do

Design and build Earthinia’s agentic AI runtime, including task routing, multi-agent orchestration, tool execution, memory, retrieval, integrations, permissions, and human approval workflows. Establish reliable evaluation, recovery, tracing, security, and performance practices while shaping the platform’s technical direction with the founder.

Requirements

Candidates should have strong hands-on experience building production AI agents and reliable systems involving LLM APIs, tool selection, orchestration, memory, external integrations, and failure recovery. They should be comfortable designing maintainable Python systems, making architectural decisions amid uncertainty, and taking substantial ownership in an early-stage startup.

Benefits

• Equity

Listed skills

• Kubernetes · Preferred
• Redis · Preferred
• PostgreSQL · Preferred
• Docker · Preferred
• Amazon Web Services · Preferred
• prompt engineering · Preferred
• Python · Preferred

Other relevant skills

Identified from the job description. Confirm important requirements above.

• Python
• FastAPI
• Agentic AI
• LLM APIs
• Tool Calling
• Multi-Agent Systems
• MCP
• Retrieval-Augmented Generation
• Vector Search
• Prompt Engineering
• Context Management
• PostgreSQL
• Redis
• AWS
• Docker
• Kubernetes

Job areas

• Technology
• Software
• Engineering
• Data & Analytics
• Management & Leadership

More jobs from Earthinia Company Services