Kaishi is currently working with a very early stage AI tech startup here in Singapore to seek an engineer who is deep in applied AI but broad enough to jump into any layer of the stack when the startup roller-coaster demands it.

How you'll spend your time:

• 80% - AI/ML: You'll get the chance to own a large chunk of our decision-making engine, working on everything from LLM prompt-/tool-design, retrieval pipelines, evaluation harnesses, agent orchestration, latency + cost tuning
• 20% - Product/full-stack: You are able to also make sure that the impact of your core work is actually getting delivered to our customers, this might include building React flows, Go APIs, observability dashboards, integrations with customer systems or other providers

WHAT YOU SHOULD HAVE

Must-have:

• 2+ yrs shipping production software, incl. any experience building with LLMs (OpenAI, Claude, etc.) or non-LLM ML engineering
• Can design and measure agent workflows: prompt chaining, RAG, vector stores, guardrails, eval suites
• You thrive when you have autonomy, own as many of the details as possible, and project manage your own work
• Willing to work in-person in Singapore, and you're serious about building a culture we'll all love

Nice-to-have:

• Experience with Twilio APIs or other telephony experience
• Experience with email delivery / SMTP systems

TECH

• Frontend: TypeScript, React, Next.js, TailwindCSS
• Backend: Go, PostgreSQL