Phosphene team

A Different Kind of AI Partner

We started Phosphene because we believed Hong Kong's businesses deserved AI built around their actual needs — not around what's fashionable in AI circles.

← Back to Home

Founded on Clarity, Not Hype

Phosphene was founded in Sheung Wan in 2016 by a small group of data scientists and systems thinkers who had grown tired of watching AI projects fail because of overpromising. The pattern was familiar: grand ambitions, rushed timelines, misaligned expectations — and teams left no better off than before.

We took a different approach. Before any project begins, we spend time understanding your organisation's actual situation — your data, your team, your processes, and the specific outcome you're hoping to achieve. Only then do we discuss what AI can realistically contribute.

Over the years, this careful, measured style has attracted clients across Hong Kong's retail, logistics, fintech, and media sectors. We've remained small by intention. Every engagement receives senior-level attention. We don't scale by diluting the quality of our thinking.

The name Phosphene refers to the light patterns you perceive when the eyes are closed — shapes that emerge from within perception itself. We chose it to represent what AI does at its best: surfaces patterns that were already present in your data, making them visible so your team can act on them.

2016

Year Founded

60+

Organisations Supported

100%

Engagements with clear scope agreed upfront

HK

Based & focused

3

Focused disciplines

What We Stand For

Honest Assessment

We tell clients what AI can realistically do for them — including when it isn't the right tool. Candid advice serves everyone better than a signed contract followed by disappointment.

Human Authority

AI in our systems surfaces options and analysis. People make decisions. We design every system so that your team remains in control — not just nominally, but practically.

Rigorous Curiosity

Every engagement starts with questions. We want to understand your data, your processes, and the context before recommending any approach. Good AI work is, first and foremost, good analytical work.

Local Grounding

Hong Kong's business culture, regulatory landscape, and cultural nuances matter. We build with that context embedded — not applied as an afterthought.

Data Stewardship

Your data is used only for your project, handled under strict confidentiality, and managed with respect for both privacy and security obligations.

Focused Scope

We offer three services because those are the three areas where we've developed genuine depth. We don't expand our scope to chase revenue — we deepen our expertise to serve clients well.

Senior Practitioners, Not Junior Generalists

Every engagement is led by someone with deep experience. We're a small team intentionally — because quality of thinking matters more than scale of headcount.

LW

Li Wei

Co-Founder & AI Systems Lead

Former data scientist at a major Hong Kong logistics firm, Li Wei spent over a decade building demand forecasting systems before co-founding Phosphene.

AC

Ananya Chawla

Co-Founder & Strategy Director

With a background in organisational strategy and applied machine learning, Ananya leads client scoping, diagnostic work, and AI readiness assessments.

CK

Chen Kai

Senior ML Engineer

Chen Kai focuses on the technical architecture of Phosphene's moderation and inventory systems, with particular expertise in multi-modal content analysis.

How We Maintain Quality

Data Security

All client data is handled under strict confidentiality agreements with end-to-end encryption and access controls.

Model Audits

Each AI system we deliver is subject to internal review for fairness, accuracy, and alignment with stated objectives before deployment.

PDPO Compliance

Our data practices comply with Hong Kong's Personal Data (Privacy) Ordinance and relevant cross-border transfer requirements.

Ongoing Review

We schedule periodic performance reviews for all deployed systems to assess drift, accuracy, and whether the model continues to serve its purpose.

AI Depth Built Over Eight Years in Hong Kong

Phosphene's work spans three interconnected disciplines: demand forecasting and inventory modelling, content policy enforcement at scale, and organisational AI readiness assessment. Each discipline draws on the same core capabilities — statistical modelling, machine learning architecture, and a thorough understanding of how human workflows intersect with automated systems.

Our Hong Kong base is not incidental. The city's retail dynamics, bilingual content environments, and cross-border supply complexity are woven into how we build. When we develop an inventory model for a Hong Kong retailer, we account for Golden Week demand spikes, typhoon signal disruptions, and supplier lead time variability from Mainland China. When we calibrate a content moderation system, we work with the specific linguistic and cultural context of Hong Kong's platforms.

We maintain active engagement with developments in responsible AI, keeping pace with guidance from the Office of the Privacy Commissioner for Personal Data and monitoring international frameworks that influence best practices. This awareness is reflected in the governance and audit processes built into every system we deliver.

Work With a Team That Thinks Carefully

Reach out to start a conversation. No commitment required — just a thoughtful discussion about what AI might offer your organisation.

Get in Touch