Ferrotype · Our Story
We built this practice around one belief: AI should serve the business, not the other way around.
← Back to HomeWho We Are
From Kwun Tong, working across Hong Kong and beyond
Ferrotype was founded in Kwun Tong by a group of engineers and analysts who had watched too many AI projects fail — not because the technology was wrong, but because the work began before anyone fully understood the problem.
The name reflects that approach. In photography, ferrotype is a direct, close-contact process — no intermediary, no approximation. The image is what it is. We carry that same orientation into AI work: direct engagement with your processes, your data, your people, and your constraints.
Since starting out, we have helped organisations across logistics, professional services, financial operations, and manufacturing in Hong Kong use AI in ways that fit their actual circumstances. We don't advocate for the most sophisticated solution — we advocate for the appropriate one.
Our team combines backgrounds in machine learning engineering, operational research, and enterprise software. We're based at 10 Shing Yip Street in Kwun Tong and work with clients across Hong Kong and the region.
7+
Years of Practice
80+
Organisations Served
3
Core Services
HK
Primary Focus
Mission & Values
What we hold to
Understand first
No proposal leaves our office before we've spent meaningful time with your team and your processes. Discovery is not a formality — it's the work that makes everything else useful.
Right-sized solutions
We recommend what fits the situation — not what's most technically impressive. A well-scoped system that your team can operate confidently is worth more than an elaborate one that requires constant specialist attention.
Plain speech
We work in a field full of jargon. Our commitment is to explain every technical decision in language that works for people who run businesses, not just those who build AI systems.
Our People
The team behind the work
Clara Leung
Principal Engineer
Leads multi-modal system architecture. Previously built data pipelines for logistics firms across the Pearl River Delta.
Marcus Tam
Automation Advisor
Specialises in process analysis and automation scoping. Fifteen years across professional services and financial operations in Hong Kong.
Priya Wong
Conversational AI Lead
Manages chatbot maintenance engagements and NLP model improvement. Background in computational linguistics and enterprise software.
How We Work
Standards we apply to every engagement
Data Confidentiality
All client data is handled under a formal confidentiality agreement and processed in compliance with Hong Kong's Personal Data (Privacy) Ordinance.
Structured QA Process
Every system we deliver passes through a defined quality review before handover — including edge-case testing, performance benchmarking, and documentation review.
Written Deliverables
Findings, recommendations, and system documentation are produced in written form — so your team has a clear record of decisions and rationale.
Iterative Review Cycles
We work in structured review cycles rather than revealing finished work at the end. This surfaces misalignments early, when they're easier to correct.
Access Controls
Only team members directly involved in an engagement access client data. Internal access logs are maintained and available to clients on request.
Ongoing Development
The team invests in regular technical training. AI changes quickly; staying current is part of doing this work responsibly.
Expertise & Context
AI practice rooted in Hong Kong's commercial reality
Hong Kong's business environment has particular characteristics that shape how AI work needs to be approached. Multi-lingual operations, data sensitivity in regulated sectors, cross-border data flows with Mainland China, and a workforce that often spans multiple functions — these factors affect what a useful AI system looks like and how it must be built.
Ferrotype's practice has developed within this context. Our workflow advisory work reflects familiarity with how Hong Kong organisations actually run — the mix of legacy systems, the operational pressures in logistics and trade, the compliance considerations in financial services. We bring that context to every engagement rather than applying a generic framework.
Multi-modal AI development requires a clear understanding of what data is actually available, how it is collected, and where gaps exist. In many organisations, the gap between what data is theoretically available and what can practically be used for AI is substantial. We assess this honestly and advise accordingly.
Chatbot maintenance is often underestimated as a discipline. A conversational AI system reflects the state of a business at the moment it was trained. As products change, processes evolve, and customer queries shift, the system drifts from usefulness without active attention. Our maintenance service treats this as the steady ongoing work it actually is.
Work With Us
Curious about what Ferrotype could do for your organisation?
We're happy to have a preliminary conversation with no expectation attached. Tell us a little about your situation and we'll respond thoughtfully.
Get in Touch