Discover Merantara

A conversation that switches between languages, a road in changing light, or someone picking up and placing an object can all provide data for AI to learn from. We organise data work for large multimodal models, autonomous driving and embodied AI so that these details are recorded clearly, described accurately and checked carefully.

A speech project, for example, starts by defining the languages, situations and recording conditions required. Participants then make recordings, which are organised, annotated and checked against the agreed requirements before use in model training or testing. Any issues are documented and addressed through corrections or new recordings, with a record of the changes.

Data collection means recording real-world situations to agreed requirements. Annotation adds descriptions, categories or other agreed information. Quality assurance checks that the data is accurate, complete and suitable for the task. Data infrastructure brings together the tools, equipment, processes and people that allow this work to continue reliably. These are the capabilities we are building.

Reliable data work depends on many kinds of expertise. Understanding language and context, organising collection, checking quality, developing tools, coordinating projects and supporting day-to-day operations all need people with different backgrounds. Observation, judgement, patience and clear communication matter throughout.

We want everyone to contribute their strengths and understand how their work fits with others' contributions, bringing different areas of expertise together to support projects and the team.

AI needs to understand different languages, cultures and ways of life to serve more people. Malaysia's diversity offers valuable experience that deserves to be explored and understood.

We want local talent to help shape new technology as well as learn to use it. Knowledge of everyday life can contribute to AI development, while project work can build lasting skills and open a path into the global AI industry. As the business develops, we plan to expand into other regions and build experience working across locations and cultures.

As an early-stage team, we are still developing our ways of working. New graduates can apply what they have learned to real problems, while experienced professionals can help refine our methods. Everyone can contribute ideas, test them in practice and help build the team.

Our team includes people with project experience in driver assistance and autonomous driving, smart home technology and embodied AI. We aim to share that experience through practical examples, documentation, collaboration and mentoring. As the business grows, project work will also help the team become familiar with the quality, documentation and delivery requirements of international projects.

Responsibilities and development opportunities will grow with individual skills, contributions, the authority assigned to each role and business needs. Training will support the responsibilities people take on. See the role descriptions for specific development programmes and their conditions.

We serve the AI industry and are building AI into our own ways of working. Whether reviewing research, analysing spreadsheets or preparing communication materials, we encourage people to choose tools that suit the task. The aim is to spend more time understanding problems, exercising judgement and finding solutions together.

What matters is a clear goal, verified results and a real improvement in the work. Not every role requires programming. We value a willingness to try new approaches, keep learning and recognise a tool's limitations. Test methods on a small scale before using them more widely. Respect access controls and confidentiality, and retain professional judgement and responsibility when handling sensitive information or making important decisions.

• Communicate clearly and solve problems together. People from different disciplines, languages and backgrounds bring different perspectives to a problem. Explain requirements and reasoning, listen to other views, raise issues promptly and work towards a solution, taking responsibility for your part.
• Work carefully and share what you learn. Pay attention to quality and detail. Share what worked and what did not so that colleagues and partners can learn from each other.
• Stay curious and test ideas. With a new tool or approach, first identify the problem it should solve. Try it on a sample task, check the outcome and adjust your approach in light of the evidence.

These principles apply to our partnerships as well as our team. We plan to bring employees and partners together for tool demonstrations, project discussions, case reviews and exchanges across disciplines. Working together on real tasks can build skills and create opportunities for lasting collaboration and shared growth.

The name Merantara combines Meranti and Antara. Meranti draws on the image of Malaysia's tropical trees: local roots, talent development, professional expertise and sustained growth. It also reflects our intention to stay connected to local communities.

Antara means “between” in Malay. In our name, it represents connections between local talent and international opportunities, cultural diversity and technology, and the real world and AI.

We want to build a company rooted in Malaysia, guided by curiosity and continuous learning. Here, local knowledge can make a difference, people can grow through practice, and careful work can help everyone progress.

View Open Roles