What are we looking for:

• 5+ years of backend engineering experience (Java / Spring preferred; strong Python or Node.js engineers are welcome)

• Hands-on experience building LLM powered applications using OpenAI, Anthropic, MCP tool orchestration, etc

• Strong prompt engineering skills, especially for structured outputs (JSON schemas, function calling, tool use)

• Experience with event-driven architectures: webhooks, message queues, async/background processing

• Comfortable operating in early-stage environments- ambiguity, speed, and ownership are part of the job (We currently support 1,800 farmers and are building toward 10M.

Good to know:

• Yes, you will go out into the field and get your feet dirty with some rice farming

• Everyone is generally remote so you need to be able to communicate really well and use many types of communication mediums including chat, call, video, etc.

• Experience in AgTech or field-operations / on-ground workforce software

• Exposure to geospatial data (GPS, mapping, routing or optimisation problems)

• Experience building multilingual LLM applications, especially Indonesian and/or Vietnamese

What You will Own:

• Design and build the Agent Control Plane (execution framework, decision logging, override system)

• Implement agents with LLM-based decision making

• Create APIs for Territory Managers to review/approve agent proposals

• Work directly with product lead on prompt engineering

How to Apply (and be seen):

AI Portfolio (REQUIRED)

You MUST demonstrate hands‑on AI experience through a portfolio showing your projects (Github Links)

• Programming Languages: Java with the Spring framework is preferred, but the team is also open to strong engineers in Python or Node.js.

• AI/LLM Application Building: Hands-on experience developing applications powered by LLMs such as OpenAI, Anthropic, or similar APIs.

• Prompt Engineering: Strong skills in creating prompts, specifically for structured outputs like JSON schemas, function calling, and tool use.

• Architectural Knowledge: Experience with event-driven architectures, including webhooks, message queues, and asynchronous or background processing.
• Operating Mindset: Comfort working in early-stage, fast-paced environments characterized by ambiguity and high ownership