prima chan.

Return to the suitcase

PRODUCT · RESEARCH · AUTOMATION

Based in Hong Kong

Hi, I’m Prima.
I work on digital products.

I turn user research into product requirements, explore AI workflows, and build automation that helps people get things done.

Curious about what AI makes possible.
Inspired by conversations, photography,
and life beyond the screen.

Interested in startups and scaleups worldwide.

Hong Kong · UK · France · Singapore

01 / SELECTED WORK

Questions worth working on.

Product work, AI workflows,
and questions behind the decisions.

AI FEEDBACK WORKFLOWS

Connecting user signals
to action with AI agents.

Feedback arrived through forms, Slack, email and the platform. I built connected AI workflows to bring repeated reports, usage signals and prior decisions together before deciding what to change.

Key decisions

  • Separated triage, investigation and refinement into three agents, each with a defined handoff.
  • Checked reports against internal records, usage signals and previous decisions before preparing revisions.
  • Built subject-specific Claude skills: the Economics pipeline checked theory accuracy, while the Geography pipeline checked information daily to keep it current.
  • Kept evidence assessment, fact-checking and approval with me.

Results

  • Reviewed all 429 topic articles across Economics (158) and Geography (271) over two months, with no factual corrections needed after review.
  • Connected grouped feedback, supporting evidence and revised drafts in one review process.
  • Published guides recorded 4,400+ student reading hours in 30 days. This engagement figure does not isolate the effect of the agents.

Reflection

An individual report rarely supplied enough context for a decision. I needed to check it against usage and previous work before acting.

My contribution
  • Connected feedback sources to a scheduled agent that filters, groups and deduplicates reports into a reviewable draft.
  • Prompted searches of internal records and feature-management systems, using SQL where needed. For feature ideas, included competitor research and prior decisions.
  • Built Claude skills to guide structure and style, and generated images to brand requirements.
  • Reviewed the exact passages flagged by audits, checked the claims and guided the next revision.
THREE-AGENT FEEDBACK SYSTEMILLUSTRATIVE

Forms · Slack · Email · Platform feedback

My review guides the next revision.

Select a stage to see the trigger, checks and handoff.

COVESTRO · USER RESEARCH & AUTOMATION

Saving an estimated 380 hours a year
on customer data updates.

Missing or incorrect customer data left tickets unassigned and blocked deliveries across APAC. I redesigned the data quality team’s email-to-SAP C4C workflow to structure requests before updating customer records.

Key decisions

  • Built a new workflow rather than extend an unstable RPA process and risk carrying its failures forward.
  • Used Power Automate to extract email data into a central spreadsheet, then UiPath to enter the structured data into SAP C4C.

Results

  • Saved an estimated 380 manual hours annually for this workflow, based on about 44 customer IDs per working day, two minutes of manual processing per ID and 260 working days per year. This annualised estimate excludes any unmeasured ongoing review or exception-handling time.
  • Raised customer-record accuracy from approximately 80% to 100% in inspected records after refinement, measured as error-free records divided by all records inspected by customer specialists.
  • Reduced runs requiring manual fixes from approximately 5% in alpha testing within the RPA team to 0% in beta testing with Hong Kong office colleagues.

Reflection

The first problem was how customer data entered the process. Structuring email requests upstream made the SAP C4C handoff more reliable.

My contribution
  • Interviewed users and mapped where missing or incorrect customer data interrupted tickets and deliveries.
  • Wrote PRDs and process flows to align requirements with stakeholders.
  • Coordinated weekly development and testing check-ins, then supported alpha and beta testing.
  • Created internal guides to support adoption.

Across five automation solutions, my CV records 2,300+ manual hours saved annually.

WORKFLOW / 01ILLUSTRATIVE

DATA PREPARATION / CLOUD

  1. 01
    Incoming emailCustomer data arrives
  2. 02
    Power AutomateExtract and structure email data
  3. 03
    Central spreadsheetStore the structured data

SYSTEM AUTOMATION / RPA

  1. 04
    UiPathRetrieve data and enter it into SAP C4C
  2. 05
    SAP C4CCustomer record updated

Reconstructed from the project workflow.

WARWICK · FUNDED CUSTOMER SEGMENTATION RESEARCH

Investigating when customer
segments should be distinct.

Segmentation shapes how businesses treat customers, but assigning each customer to one group can hide behaviour shared with other groups. I investigated when distinct segments are justified across transaction and subscription datasets.

Key decisions

  • Compared K-means and Fuzzy C-means rather than choose a method upfront.
  • Used RFM measures as a proxy for behaviour and adapted them for subscription data.
  • Evaluated K-means with silhouette score and Fuzzy C-means with the fuzzy partition coefficient.

Results

  • Proposed a framework: assess behavioural overlap with Fuzzy C-means first; consider K-means when overlap is low and retain soft clustering when it is high.
  • Found more overlap in the subscription dataset and clearer groups in the transaction dataset, reversing my expectations.
  • Secured £1,500 in competitive Warwick URSS funding and presented the study at the URSS showcase.

Reflection

Customer behaviour should guide the choice of segmentation method. The proposed framework needs to be considered alongside cluster-quality checks and practical constraints; my findings cover two datasets and two methods.

October 2024–November 2025 · University of Warwick

My contribution
  • Secured funding and took the study from proposal to public display at Warwick’s URSS showcase.
  • With supervision, used CRISP-DM to compare K-means and Fuzzy C-means across transaction and subscription datasets.
  • Adapted RFM measures for subscription behaviour and compared the overlap in each setting.
  • Used the findings to propose checking behavioural overlap before choosing a segmentation method.
A simplified version of my proposed framework

Assess behavioural overlap with Fuzzy C-means

Less overlapConsider K-means

Distinct groups may support uniform treatment.

More overlapRetain Fuzzy C-means

Allow partial membership across groups.

Use alongside cluster-quality checks and practical constraints. Findings are limited to the two datasets and methods examined.

12345This study
  1. 1Hard vs. soft clustering
  2. 2Customer segmentation
  3. 3Business models
  4. 4Uniform treatment assumptions
  5. 5Technical evaluation
EXPLORE THE IDEAILLUSTRATIVE
Hard clusteringK-means

One group per customer.

Soft clusteringFuzzy C-means

Membership across groups.

The same illustrative customers appear in both plots. The slider moves the groups closer together.

Conceptual comparison · synthetic data.

02 / HOW I WORK

Frame, write, prototype, refine.

Select a stage to explore.

Learning feeds back into framing and PRDs.

Skills I bring to the work

Discovery & research

  • User interviews & focus groups
  • Workflow mapping
  • Survey research & feedback analysis
  • Microsoft Forms
  • Tally

Product definition

  • PRDs & user stories
  • Acceptance criteria
  • Alpha, beta & user acceptance testing
  • Notion
  • Jira
  • Confluence

Prototyping & AI workflows

  • Interactive prototyping
  • Agentic workflows
  • Human review & iteration
  • Figma
  • Claude
  • Claude Code
  • Codex

Analytics & automation

  • Customer segmentation · CRISP-DM
  • Dashboard design & KPI analysis
  • Process automation
  • Python
  • PostgreSQL
  • Tableau
  • Power Automate
  • UiPath
  • Excel VBA

03 / EXPERIENCE

Across products and contexts.

Product, research,
and operational leadership.

General Learning YC F24

Product Manager

Product management in digital education.

June–November 2026

Covestro

Customer Excellence & Digital Solutions
Placement Intern

Interviewed 15+ users across six APAC markets, documented requirements, and developed automation solutions.

Sep 2023–Aug 2024

Jardine Matheson

Summer Intern, Learning & Development
Group Human Resources

Moderated two focus groups with 30+ front-line engineers and analysed feedback from 500+ APAC employees across nine Learning & Development sessions.

Jun–Aug 2023

Juvenate Consulting

Managing Director

Led a team of five to redesign project delivery and student consultant recruitment.

Jul 2022–Dec 2024

Request my CV View LinkedIn

04 / A LITTLE MORE ABOUT ME

A curious eye.
A perspective of my own.

Prima Chan taking a self-portrait on a railway platform

I’m based in Hong Kong, with experience living and studying in the UK, France, and Singapore.

Moving between these places has made me curious about how people experience the world differently. My mission is to help people understand one another across cultures, through the stories I tell and the products I help build.

I like understanding how feedback, decisions and delivery connect, then building a process around them.

I’m interested in why a process breaks down and whether its assumptions hold. At Covestro, I interviewed users across six APAC markets and learnt to build automation. At Warwick, I secured funding to investigate when customer segments should be distinct .

Outside work, I explore different perspectives through film photography and learning Korean and Thai. Conversations on my travels have shaped what I notice and how I photograph people respectfully.

NUS Full Stack Development with AI Executive Programme · 2026–2027 · Ongoing

INSEAD Master in Management · Class of 2027

WARWICK BSc Management with Finance

A person sitting outside a blue shutter beneath a striped awning, photographed by Prima Chan A coffee sign and building facade, photographed by Prima Chan
A small lighthouse on a grassy hill, photographed by Prima Chan A white van beside a garden and buildings, photographed by Prima Chan
FROM MY FILM ARCHIVEPhotographs by Prima Chan · whcamite
Explore my photography

CONVERSATIONS / SINCE 2022

Sharing what I’ve learnt.

Since 2022, I’ve been invited back to my high school to share my university experiences with current students. I want these conversations to help students see new possibilities and feel more confident exploring their own path.

Alumni sharing · NPL class of 2021

Prima with fellow alumni, staff and students at her high school sharing event
An alumni sharing session at my high school in 2026.

WHAT I WANT TO KEEP MAKING ROOM FOR

Human creativity,
alongside AI.

I’m excited by what AI makes possible, and I believe human creativity deserves room to grow alongside it.

Photography gives me a reason to slow down and notice. I want those observations and conversations with people to keep shaping the questions I bring to product work.

05 / SAY HELLO

Let’s exchange ideas.

I’m interested in startups and scaleups worldwide.
If you’d like to discuss a role or exchange ideas, I’d be happy to hear from you.

Email me