Engineering
Applied AI Engineer
Singapore or remoteFull-time · Graduates welcome
Who we are
Synscribe builds the platform that carries the work behind a marketing strategy. It digs through the detail, finds the openings, drafts and publishes the work, and tracks what moved.
Agencies use it to take on more clients. Founders and CEOs use it to run lean. B2B teams use it to get found on Google and quoted by ChatGPT.
We are a technology company in the SEO and GEO industry. We write our playbooks, use them in real client work, and build the parts that work into our product.
Our AI work is strictly applied. We do not run a research lab for its own sake. We study a method because it can improve a real user result, then we test it and decide whether to use it.
What you'll do
You will find useful developments in AI and turn them into reliable product and team systems.
You will move between source code, focused technical tests, product work, and user conversations. You will connect what a framework claims to do with what happens when a real person uses it.
You will:
- Study new models, agent frameworks, harnesses, evaluation tools, and engineering methods.
- Go beyond product pages and documentation. Read source code when you need to understand how a system really works.
- Run focused spikes to answer one technical or product question at a time.
- Compare a new method with the current method. Test quality, failures, speed, cost, and user effort where they matter.
- Recommend a clear next step: use it, change it, watch it, or stop.
- Build and improve agent workflows, tool connections, context handling, evaluations, guardrails, logs, and monitoring.
- Turn useful experiments into reliable features or internal systems.
- Work with the delivery team to find repeated problems that Product can solve once for many users.
- Get in front of users. Watch them use the product, ask where they get stuck, and connect their real work to the technical decisions that you make.
- Explain technical choices and trade-offs to engineers and non-engineers without hiding the important nuance.
This role does not prescribe one programming language, database, retrieval design, or agent framework. We want a strong engineer who will inspect the current system before proposing a replacement.
What we're looking for
You are a good engineer with deep technical curiosity. You do not stop when an abstraction becomes unclear. You trace it into the source code, form a clear question, and build the smallest useful test.
You can be a good fit if:
- You have strong software-engineering fundamentals and can show a working system that you built.
- You can explain how a system works, where it can fail, and why you chose one trade-off over another.
- You enjoy reading unfamiliar code and following a problem beyond the documentation.
- You are extremely curious, but you can keep that curiosity focused on a clear commercial or user result.
- You can run a short spike with a clear question, baseline, test, and decision.
- You care about evaluation, data quality, reliability, security, cost, and user effort.
- You want to speak with users instead of making every product decision from behind a screen.
- You can teach the team how to use a system and check if the system improved the real work.
- You can show what happened after the demo, including failures and changes.
Fresh graduates are welcome. Graduate research, open-source work, internships, and personal projects are useful evidence when they solve or test a real problem. A paper or benchmark without a path to practical use is not enough for this role.
What to send us
Send your CV and one or two examples of systems that you built.
For each example, tell us:
- the problem and the user;
- how the system worked;
- what part you owned;
- how you tested it;
- one failure or unexpected result; and
- what happened after the first version.
A small system with clear reasoning and good tests is stronger proof than a large AI wrapper with no real users or useful measurements.