ELife Transfer
Website:
elifelimo.com
Job details:
Job Title: Forward Deployed AI Engineer
Reporting to: Head of AI & Agentic Automation
Embedded with: Elife's engineering squads and internal operations teams (CE, Supplier Ops, Dispatch)
About Elife
Elife is the Enterprise Super App Enabler — the global B2B infrastructure powering rides and instant delivery for the world's largest enterprise platforms. Through API, SDK, AI Agentic, and White-Label integration, Elife connects 100+ enterprise apps — super apps, fintech platforms, OTAs, airlines, map platforms, ride-hailing and delivery apps — to a network of 100+ ride suppliers, 70,000+ local fleets, and 100+ delivery partners across 182 countries.
About the Role
Elife is building "The Brain" — a centralized AI nervous system that connects LLM capabilities directly into our pricing engine, dispatch backend, and driver tracking systems. The Head of AI is architecting that brain. You are the one who deploys it.
This is not a research role and it is not a platform role. You will spend most of your time sitting next to the people who will actually use what you build: our backend engineers wiring up tool-callable APIs, our frontend developers integrating AI copilots, and our operations teammates who need agentic workflows to replace manual processes. Your job is to understand their problems better than they do, ship working AI solutions inside their environment, and make sure those solutions actually get adopted.
If you like writing code that ships on Friday and gets used on Monday, this role is for you. If you prefer to write papers or build abstractions in isolation, it is not.
What You Will Do
- Embed directly into one or more of Elife's agile engineering squads or operational departments. Learn their codebase, their pain points, their workflows, and their definitions of "good." Become the AI expert they trust.
- Build agentic workflows, RAG pipelines, and LLM-powered tools tailored to specific internal use cases — for example: an agent that triages dispatch anomalies, a copilot that helps backend engineers write tool-calling endpoints, an AI QA assistant that auto-generates test cases against our pricing logic, or a workflow that lets Supplier Ops onboard a new city without writing SQL.
- Wire LLMs into Elife's proprietary backends through the function-calling and tool-calling APIs that the Head of AI's team is exposing. You will be one of the heaviest consumers of those APIs and one of the loudest voices on what they need to do.
- Pair-program with traditional backend and frontend developers. Review their code. Show them how to think in prompts and orchestration rather than only in hard-coded logic. Leave every team you embed in measurably more AI-native than you found it.
- Own the full lifecycle of what you ship: prototype, evaluate, deploy, monitor, iterate. Track real adoption metrics — not demos, not slide decks. If your tool isn't being used, that's your problem to solve.
- Feed learnings back to the Head of AI and the central AI team so common patterns get promoted into shared frameworks and platform capabilities.
Who You Are
- 3+ years of hands-on engineering experience, with at least the last year spent shipping LLM-powered applications to real users. You've built things with tool-calling, RAG, agent frameworks (LangChain, LlamaIndex, LangGraph, or equivalent), and you know where they break.
- Strong general-purpose engineer first, AI specialist second. You can read an unfamiliar Python or Go codebase, find the right place to plug in, and write code that the team owning that codebase is happy to merge. Comfortable with relational databases, REST/gRPC APIs, and the basics of distributed systems.
- Customer-obsessed in the FDE sense: you'd rather sit in a dispatch ops room for two days watching people work than spend two days tuning a model in isolation. You ask "what would make your job easier" before you ask "what model should I use."
- A teacher and a translator. You can explain to a senior backend engineer why function-calling will not destroy their service, and you can explain to a non-technical operations lead what an agent can and cannot reliably do. You don't condescend in either direction.
- Comfortable with ambiguity. Internal customers rarely give clean specs. You're good at turning a vague complaint into a scoped, shippable project.
- Bias toward shipping. You'd rather deploy something imperfect this week and iterate than deploy something perfect next quarter.
Nice to Have
- Experience in mobility, logistics, marketplaces, or any domain with real-time dispatch and dynamic pricing.
- Background as a full-stack or backend engineer before moving into AI — you remember what it felt like to have an AI team throw a half-working prototype over the wall, and you've sworn never to do that to anyone.
- Experience evaluating LLM systems in production (eval harnesses, regression suites, cost/latency tradeoffs).
Click on Apply to know more.