Be Part of the Founding R&D Team | UAE-Based Company | Opportunities in Jordan

𝗥𝗲𝗺𝗼𝘁𝗲 · 𝗔𝗜 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 & 𝗣𝗹𝗮𝘁𝗳𝗼𝗿𝗺 · 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿 𝗜𝗜–𝗜𝗜𝗜, 𝗟𝟰–𝗟𝟱 · 𝗠𝗝𝗗𝟭𝟬.𝟬.𝟭

Product engineer who owns the core of the application in which 𝗽𝗲𝗼𝗽𝗹𝗲 𝘀𝘂𝗽𝗲𝗿𝘃𝗶𝘀𝗲 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀: the data that people and agents change at once, what a person sees while a result is pending, and the 𝗿𝗲𝗽𝗹𝗮𝘆 𝘁𝗵𝗮𝘁 𝘀𝗵𝗼𝘄𝘀 𝗲𝘅𝗮𝗰𝘁𝗹𝘆 𝘄𝗵𝗮𝘁 𝗵𝗮𝗽𝗽𝗲𝗻𝗲𝗱. Greenfield.

THE TECHNICAL CHALLENGE

We build agents on open-weight LLMs that operate web applications through the browser. People supervise these agents in one application, and people and agents change the same things at once. Results arrive half finished, and a person can correct, take over or stop at any moment. The engineering decides which copy of the data counts and what undo means when one participant is a machine. An agent is held to the steps it may take, and every step can be replayed afterwards. This role owns that core and builds the small simulated applications the agents are tested against. Those applications have sessions that expire halfway through a form and identifiers that change on every page load.

KEY RESPONSIBILITIES

• Own the application core: which copy of the data counts, what each participant's undo takes back, and how a person sees what is pending or unknown

• Keep an agent to the steps it is allowed to take, approval and stop among them, so that anything else is blocked before it has an effect

• Build whole features end to end, client, server and stored data

• Build and maintain the small simulated applications the agents are developed and tested against, every fault switched on or off by configuration and behaving the same way on every run

DESIRED QUALIFICATIONS

• The data layer of a shipped application in which several people work on the same thing at once, owned through its second year: syncing, working offline, undo, what happens when two changes conflict

• Merged contributions to a library for shared data or syncing, touching how changes merge or how data is stored

• A replay or simulation system built that gives the same result every time, with the order of events kept right when several changes happen at once

• A voice pipeline shipped in a web application: turn-taking, interrupting the speaker, echo cancellation, response-time limits

EXPECTED QUALIFICATIONS

• T-shaped: deep in one domain, with working breadth in a neighbouring one

• Structures a large, incompletely specified problem and drives it to a working result independently; treats two things changing at once as a design question from the start

• Complex features shipped end to end in TypeScript across client and server, with changes shown before the server confirms them, reconnects and working offline handled; uses AI coding tools daily and verifies their output

• The stored data and the updates of a shipped application owned: schema changes, migrations and the path from one version to the next

——

HOW WE WORK

• Product engineering: we own what we build and run it in production

• Small teams, two-week cycles, working software at every review

• AI coding tools are part of the standard workflow

WHAT WE OFFER

• Founding-team scope

• AI-augmented engineering environment

• Access to on-premise Nvidia B200s

• Flexible work environment

PROCESS

• Introductory call

• Technical conversation

• Practical session; the format is agreed with you

In coding exercises, AI tools are allowed and expected. No LeetCode.

WHO WE ARE

New product organisation as part of a large semi-government in Abu Dhabi. International, ex-FAANG team. Completely greenfield, with a modern tech stack.

——

REQUIREMENTS TO BE CONSIDERED

• Clear written and spoken English

• 4+ years building web applications in production; strong TypeScript; own code operated in production, with users

• Has designed the data of an application in which several people change the same thing at once, and can say what merged, what did not, and why

• Bachelor's degree in any field, or self-taught with a track record of open-source contributions

RELATED TECHNOLOGIES AND CONCEPTS

• Data that several people change at once: the order of events, positions that survive other people's edits, undo and redo with several participants

• State handling: XState, Zustand or Jotai

• Storage: SQLite and IndexedDB

• Real time and voice: WebSockets, WebRTC, LiveKit Agents or Pipecat, voice activity and turn detection, speech-to-text and text-to-speech

• Test applications: Playwright, seeds that make a run repeatable, faults switched by configuration