Nov 2026 - Present
- Product owner for wealth and trading inside the main Revolut app.
Wealth & tradingTechnical product
James Liu, product manager for wealth and trading at Revolut: trading products, with AI.
I'm James, a product manager who started out as an engineer. I work on wealth and trading at Revolut, after leading derivative products at Crypto.com and building AI products at ByteDance. On the side I build AI agents and video tools.
- Product manager
- 4 yrs
- Founder
- 6 yrs
- $800K cumulative ARR
Experience with
- Revolut
- Crypto.com
- ByteDance
- Trendsi
- Ernst & Young
Professional journey
Experience across trading, derivatives and applied AI.
Four years as a product manager, from AI at ByteDance to trading products for 140 million users, after starting out as an engineer. Six years as a founder alongside, with $800K in cumulative ARR. The through-line is shipping things that move a number.
Sep 2025 - Nov 2026
+60%
Prediction Markets MAU, QoQ
+90%
New-user trading volume, referrals
−90%
AI workflow turnaround
- AI and trading for 140 million users worldwide. PM for the main app's trading surface and the infrastructure beneath it, focused on derivatives: options, futures and prediction markets.
- Prediction Markets: owned market mechanics, pricing and quotation logic, market operations and rewards. Grew monthly active users 60% quarter on quarter with market launches and lifecycle campaigns tied to the FIFA World Cup and NBA Finals.
- Worked with engineering on an AI trading agent that lets users place and manage trades in natural language through OpenClaw and third-party bot integrations: the tool-calling surface, the context passed to the model and the safety constraints. Contributed code via PRs.
- Designed and automated the team's internal AI workflows (RAG retrieval, multi-agent orchestration, PRD drafting and working prototypes hosted on the company domain), cutting turnaround on that work by 90%. Built on Claude Code, MCP and the Google Workspace CLI.
- Spearheaded the main referral programme and the welcome bonus, lifting new-user trading volume 90% and 15% respectively.
- Compliance for app onboarding, KYC, AML and anti-fraud, alongside the geographic expansion of trading products into the EEA and additional US states.
DerivativesPrediction marketsAI agentsGrowth
Jul 2024 - Sep 2025
$24M
Enterprise AI deal
200K+
User licences
>20%
Lead conversion, per page
- Converted to full time in 2024 and received a spot bonus in Q2 2025.
- Ran the Perplexity AI partnership end to end, from model evaluation and commercial negotiation to technical integration, closing a $24M enterprise deal covering 200,000+ user licences.
- Co-led a 16-person team building a multi-modal AI content generation platform for SEO: the vector embedding pipeline over the Lark knowledge base, the semantic retrieval layer and the generation workflow. It drove 20% of organic SEO traffic and lifted unique monthly visitors 20% in 3 months.
- Redesigned 20+ core high-traffic web pages, lifting lead conversion by more than 20% per page on average, and grew ARR 7% through modular B2B SaaS referral initiatives.
Multimodal AIRAGAI partnershipsB2B SaaS
Aug 2023 - Jul 2024
+35%
Cart conversion
−90%
Fraudulent txns
$20K
Saved monthly
- Raised cart conversion 35% by leading development of a product similarity recommendation engine at the #1-ranked Shopify and TikTok App Store startup (Series A).
- Cut fraudulent transactions 90%, saving about US$20,000 a month, with Stripe 3D Secure and Radar risk rules.
RecommendationsPaymentsFraud
May 2023 - Aug 2023
1M+
Users served / yr
- Led frontend development of a real-time analytics dashboard for an international enterprise client serving 1M+ users a year.
ReactJS.NETC#Azure
Apr 2021 - Present
S$200K+
Annual revenue
30,000+
Customers
- Grew a DTC wellness brand to S$200K+ annual revenue and 30,000+ customers, running acquisition on AI-generated video ads and UGC alongside lifecycle and referral campaigns.
- Built its growth stack: PayNow checkout, a WhatsApp reorder engine, a referral captain kit and an SEO content engine.
DTCAI video adsLifecycle
Select a role to read the detail
Selected work
Things I've built, and what they proved.
An AI trading desk that grades its own agents, two market-design simulators, an AI video studio, and the DTC brand those videos sell for.
AI agents · Trading01Quorum
A crypto trading desk where nine AI agents argue, a deterministic risk kernel decides, and I approve every order. Built to test the one claim four viral AI-trading posts never measured: that the AI adds anything.
9 strategy runs · 0 survived the Lab
Multi-agent LLMsPythonReactBacktesting+1
Market design · Prediction markets02Market Design Toolkit
Two tools that turn market-quality arguments into budget lines. Market Depth Budget prices “this market feels dead” with Hanson's LMSR; Rewards Check runs Polymarket's published maker-rewards rule and shows what the pool buys per filled share.
0.5% slippage bar: $34,484 subsidy per market
LMSRMarket microstructureJavaScriptSVG charts
Multimodal AI · Video03Kino
An AI image and video studio for small brands. Guided tasks preview the result before anything is made, an ad maker writes and compliance-checks TikTok and Meta ads, and Advanced runs the official ComfyUI workflows for MiniMax H3, Qwen Image 2.1 and LTX-2.3 on RunPod GPUs.
Live web demo · 8 guided tasks
ComfyUIRunPodMiniMax H3Qwen Image 2.1+2
DTC · AI video ads04HerbalBath
The DTC wellness brand I co-founded, grown on AI-generated video ads and UGC. I build its growth stack too: PayNow checkout, a WhatsApp reorder engine, a referral captain kit and an SEO content engine.
S$200K+ annual revenue · 30,000+ customers
AI video adsUGCNext.jsWhatsApp Cloud API+1
Toolkit
What I know, from markets to the models underneath.
Three domains I've owned and three I build with. I started as an engineer and moved into product, so I work the whole path: the spec, the PR, the eval and the dashboard that says whether it worked.
Trading & Derivatives
- Options & strike options
- Perpetual futures
- Prediction markets
- Tokenized stocks
- Liquidity & market making
Contract mechanics, pricing and the plumbing under the trade button.
Growth & Rewards
- Referral programmes
- Welcome bonuses
- Incentive design
- KYC-to-first-trade funnels
- A/B testing
Getting the first trade to happen, then the second.
Risk & Compliance
- KYC & AML
- Licensing & market expansion
- Stripe 3DS & Radar
- Risk limits & kill switches
- Safety constraints for agents
The unglamorous work that decides which markets you can launch in.
AI Agents & Evals
- Tool-calling design
- Context & memory
- Multi-agent systems
- Evals & calibration
- MCP integrations
Agents that touch money, graded against plain rules.
Multimodal & Video AI
- Image & video generation
- ComfyUI workflows
- FLUX, Qwen-Image, Wan 2.2
- RAG & vector search
- AI video ads & UGC
From a structured brief to a captioned 9:16 ad.
Build & Data
- Python & SQL
- TypeScript, React & Next.js
- Claude Code & Cursor
- AWS & REST APIs
- Amplitude & Mixpanel
Enough depth to write the spec, read the PR and ship the prototype.
Academic
Four years of computer science, and the detours that mattered most.
Honours in Computer Science at NUS, graduate classes at Stanford during a year in Silicon Valley, and a habit of entering things I wasn't sure I could win.
2021 - 2025
National University of Singapore
Bachelor of Computing, Computer Science
Honours (Distinction)
Singapore
2023 - 2024
Stanford University
Graduate coursework, Management Science & Engineering
Graduate classes in AI, computer science and entrepreneurship, during the Silicon Valley year with NUS Overseas Colleges.
Stanford, California
2024
NUS Venture Initiation Programme
Won a $10,000 grant for a technical startup from NUS Enterprise.
National University of Singapore
2023 - 2024
NUS Overseas Colleges, Silicon Valley
A year-long tech internship programme in San Francisco while studying at Stanford University.
National University of Singapore
2023
2nd Place, Stanford Hackathon
Placed 2nd among 100+ participants at the Stanford Startup Society hackathon.
Stanford Startup Society
2023
Student Mentor, Software Engineering
Finished in the top 5% of the NUS cohort for CS2103T Software Engineering, then mentored the next cohort.
NUS School of Computing
Also
- School of Computing Ambassador2022 - 2025
- School of Computing Leadership Programme2022 - 2024
- Orbital (Independent Software Development)2022
- Citadel Datathon2023
Activities
NUS Kent Ridge HallShopee Code League '22Hack4Good '23
Get in touch
Building something interesting? Let's talk.
I'm always up for a conversation about product, applied AI, or an idea you can't stop thinking about. I reply to everything.