AI Product Manager & Builder working across consumer and enterprise AI, with a focus on agents, multimodal workflows, and evaluation.

  • AI PRODUCT BUILDER
  • +
  • STATISTICS × FILM
  • +
  • CHINA ↔ U.S.
NAJIA PAN
▧

Systems thinking

Design clear boundaries, handoffs and validation for AI and people.

✳

Product creativity

Turn real user needs and AI’s capabilities into useful experiences.

◒

Design taste

Unite visual storytelling, brand and interaction with product logic.

With AI, I reach further.
Imagination gives possibility form; human judgment gives it direction.

Selected Works

01–03 / 05
ACT I / LEMON BOARD GAMES

Lemon.
A world worth playing.

Starting with the stores I co-run, I bring brand visuals, a customer–staff mini program and AI assistants into everyday operations.

EXPERIENCE BUILD · INTERNAL TESTING
My role
Co-owner / Product & visual direction
DELIVERABLES
WeChat mini program (customer + staff) · AI assistant matrix: rules, Turtle Soup and strategy
Explore
Brand · UI/UX · Product & systems engineering · Evaluation · Deployment · Human–AI delivery
Enter Act I ↗
ACT II / UNSEEN

What lies
under the answer?

From confirming rules to open-ended decisions. Unseen explores decision metacognition: reframing questions, examining evidence, and keeping judgment in human hands.

EXPERIENCE BUILD · INTERNAL TESTING
My role
Independent product & system design
DELIVERABLES
Unseen AI-native web product · Lemon AI strategy assistant · Board-game rules assistant
Explore
User research · Agent matrix architecture · Multimodal recognition · Framework & evidence routing · Human–AI evaluation
Enter Act II ↗
ACT III / CREATIVE WORKFLOWS

Keep judgment
in creation.

From director workflows and generative evaluation to human visual judgment.

ENTERPRISE WORKFLOW RESEARCH & EVALUATION
MY ROLE
Enterprise AI Product Manager · Research, workflows and evaluation
DELIVERABLES
AIGC director workflow design · Multimodal human–AI evaluation
EXPLORE
Creative intent · Storyboarding & generation · Evaluation calibration · Design decisions
Enter Act III ↗
ACT IV / INTENT / SYSTEMS

Make execution
accountable.

From sales knowledge and intent routing to Data Warehouse Agent evaluation.

SEPARATE WORK CASES · SANITISED
MY ROLE
Enterprise AI Product Manager · Requirements, routing and evaluation
DELIVERABLES
AI Sales Copilot: knowledge base/RAG & intent routing · Data Warehouse Agent evaluation
EXPLORE
Knowledge governance · Intent & field routing · Failure attribution · Acceptance boundaries
Enter Act IV ↗
ACT V / EVOLVING JUDGMENT

Give judgment
a next round.

From discovery in Foodies to Learning OS and ChainThread.

Personal AI explorations across apps, workflows and hackathons.
MY ROLE
Independent product research & building · Workflows, evaluation and prototypes
DELIVERABLES
Foodies app prototype · Learning OS product-judgment workflow/Skill · ChainThread research prototype
EXPLORE
Discovery · Cognitive responsibility · Human Review · Evidence & revision
Enter Act V ↗

Experiments & fragments

Choose fragments ↗

About Me

From life in Los Angeles and Palm Springs to Yesanpo near Beijing, I have spent eight years moving between two worlds.

I am an observer—of data and people, of what technology can do and what people might come to love. UCLA taught me statistics; film taught me how to frame a shot. Numbers are my language. Images are my eyes.

I connect things that seem unrelated—data and narrative, technology and human feeling, business and warmth—and turn them into things people can use, try and put to the test.

Away from the screen, I enjoy strength training. Control of the body and a precise sense of muscular tension in space draw on the same attentiveness I bring to understanding people.

Lemon is the shop I help run: from a physical place to a digital product, from visual identity to systems, from AI assistants to evaluation. Unseen holds a question I keep returning to: when AI gives us answers, can it leave judgment in human hands?

Najia in a field of flowers

Have something
worth building?

I'd love to compare notes on AI products, systems, and the experience around them.