Leads the team this role joins
In-house, owns the opening
Put the opening up
From your own Gmail and LinkedIn.
About the role
This is a full-time, remote role for an applied machine learning engineer who evaluates and adapts language models for a phone-based AI product, open to candidates in Asia. The employer is Bjak, an insurance platform headquartered in Malaysia that serves customers across Southeast Asia. The work centers on testing open-weight models against the product's use cases and deciding where fine-tuning is needed. The product aims to make everyday messaging, notes, calendars and tasks more proactive with minimal prompting.
What you’ll do
- Assess existing foundation models such as Qwen, Llama and Gemma against the product's phone use cases.
- Decide which capabilities existing models already handle and which need adaptation.
- Run lightweight fine-tuning experiments using SFT, LoRA, QLoRA and related methods.
- Improve intent recognition, instruction following, tool calling, structured output and multi-step task execution.
- Build synthetic data and datasets for phone-specific workflows.
- Create evaluation sets and benchmarks that measure model quality and task completion.
- Compare cloud and local inference options and different modeling approaches.
- Work with the Agent Runtime team to improve model behavior in live agent workflows.
What they’re looking for
- A solid background in machine learning, deep learning or NLP.
- Hands-on experience with LLMs and open-weight models.
- Experience fine-tuning models with SFT, LoRA or QLoRA.
- Working knowledge of PyTorch and common ML tooling.
- Understanding of LLM evaluation, benchmarking and dataset preparation.
- The ability to run experiments, diagnose model failures and turn findings into measurable gains.
Nice to have
- Experience with RL or preference optimization.
- Experience with agent systems, function calling and tool-use training.
- Experience with synthetic data generation and preference datasets.
- Experience with model distillation and smaller language models.
- Experience deploying models in production environments.
What’s on offer
- A full-time, remote working arrangement.
Skills
Questions about this role
Is this role remote?
Yes. The role is remote and open to candidates in Asia.
Is it full-time?
Yes, the role is full-time.
What experience is needed?
The listing asks for practical LLM work, fine-tuning experience with SFT or LoRA/QLoRA, and PyTorch. It does not state a number of years.
Does the posting list a salary?
No. The listing does not state compensation.