T Talwen
About Talwen

AI work grounded in
genuine engineering.

Talwen was established to give Malaysian organisations access to applied AI expertise without the opacity that often surrounds large technology engagements.

Back to Home
Talwen team in Kuala Lumpur
Our Story

How Talwen came to be

Talwen was founded in Kuala Lumpur by practitioners who had spent years working on data engineering and machine learning problems inside large Malaysian corporates. What became clear over time was that the most common reason AI projects failed had nothing to do with the technology — it was a mismatch between what the business actually needed and what the technical team went off and built.

The firm was established to close that gap. Our work begins with listening and framing before any architecture decisions are made. We believe the most useful AI system is usually a modest one that does one thing reliably, rather than an ambitious platform that never quite reaches production.

Operating from Jalan Ampang, we work with operations and technology teams across sectors including financial services, logistics, media, and professional services. Our engagements are deliberate and structured — we don't take on more projects than we can support well.

The name Talwen comes from the Welsh word for "bright forehead", a reference to the sundial — an object that tells the time not by electronics but by understanding the geometry of light and shadow. We think that's a fair metaphor for good analytical work.

Our Team

The people behind the work

AZ

Ahmad Zulfikri

Founder & Principal AI Engineer

Over twelve years in enterprise data engineering across Malaysian banking and logistics sectors. Leads all process mining and AI architecture engagements.

NF

Nur Farhana

ML Engineer & Search Specialist

Specialises in embedding models and information retrieval. Responsible for the technical design of multimodal search platform engagements.

RK

Raj Kumar

Delivery Lead & Workshop Facilitator

Brings a background in programme management and technology strategy. Facilitates AI alignment workshops and oversees engagement delivery quality.

How We Work

Quality standards and professional protocols

Defined scope before work begins

Every engagement starts with a written scope document outlining deliverables, timeline, and acceptance criteria. No surprises mid-project.

Data privacy and security

Client data is handled under non-disclosure agreement as standard. We follow PDPA (Malaysia) requirements and do not retain client data beyond the engagement period.

Version-controlled delivery

All code deliverables are version-controlled and transferred to client repositories. Documentation accompanies every technical handover.

Regular progress communication

Fortnightly written progress updates for multi-week engagements. Key decisions and change requests are documented and mutually agreed before proceeding.

Tested before handover

Deliverables are validated against the agreed acceptance criteria before the final handover meeting. A structured post-handover support window is included.

Knowledge transfer included

We don't leave teams dependent on us. Handover sessions and documentation are designed so your people can maintain and extend what we build.

Applied AI expertise in Kuala Lumpur

Talwen operates at the intersection of data engineering, machine learning, and organisational process improvement. Our work focuses on the practical end of AI — the systems that process real operational data, serve real search queries, and sit inside real workflows used by real teams.

Based at 28 Jalan Ampang in Kuala Lumpur, we serve organisations across Peninsula Malaysia and are familiar with the specific technology landscape, talent market, and regulatory considerations that shape AI adoption in Malaysian businesses. We understand PDPA obligations, the SST treatment of professional services, and the procurement realities facing teams in both public-linked and private-sector organisations.

Our three service lines are deliberately distinct: process mining engagements surface operational facts hidden in system logs; multimodal search builds retrieval capability across diverse content types; and our alignment workshop reduces the planning friction that causes so many AI projects to stall before they start. Together, they represent the most common points where organisations need structured external support.

Curious whether we're a good fit?

An initial conversation costs nothing and helps both sides understand whether there's a worthwhile engagement to scope.

Start a Conversation