I turn complex client and operational needs into new product capabilities, then stay close through engineering, QA, launch and live use.
0-to-1 work: workflow execution, client-owned identity and AI-ready product memory.
Anjelika Tan
Product Manager
About me
Anjelika Tan
I'm a Product Manager who enjoys finding the bigger platform problem behind a one-off client request. I work across workflows, integrations, identity and commerce, including the operational details that determine whether a feature still works once it is live.
Capabilities
Most of my work sits across APIs, webhooks, workflows, Store, gamification and AI-enabled product operations.
API
Product logic noteI work with platform APIs from both sides: reading client documentation and payloads, then defining how our own endpoints should validate, fail and expose data. That includes permissions, type-specific date and time formats, bulk operations and product flows where an ambiguous contract can create incorrect rewards or account state.
Selected work
The lead case follows a client programme from data and product rules through launch and live operation. The next three show how I shaped and expanded the platform systems underneath it with engineering, design and QA.
Client names, screenshots and identifying operational details are generalized or omitted. Metrics are labelled as production, QA/test or bounded operational evidence.
Led the product and integration workstream for a multi-region brokerage loyalty programme, translating client goals into embedded experiences, data contracts, campaign mechanics, operational tooling and automated fulfilment.
Observed result
The programme reached live multi-region operation across embedded reward experiences and automated trading-credit fulfilment, with reconciliation and incident evidence used to tighten the product after launch.
Shaped a voucher Store into a configurable redemption platform spanning eligibility, permissions, order states, refunds, SDK and API contracts, and operational investigation.
Observed result
Production work made historical API responses return the intended statuses for nearly 3,000 purchases without manually rewriting every order.
Helped evolve a visual workflow builder into an operable execution platform with triggers, actions, transformations, concurrency, test and retry behaviour, branching, bulk operations and run history.
Observed result
Workflow paths reached production use and exposed real missed and duplicate-trigger cases that sharpened the observability and recovery model.
Turned a failing large-user list into a wider operability programme across normalized state, queued imports and updates, cancellation, API Logs, background export and export history.
Observed result
QA loaded a 300,000-record list and completed a 50,000-record multi-field update in under 15 minutes; production use is proven at smaller batch sizes.
AI practice
I use AI as a layer over product context, not a substitute for it. I built a product memory that records why a feature exists, what changed, where the evidence came from and which decisions still need a person.
Why I built it
Important context was scattered across tickets, chats, PRDs, QA notes, designs and production follow-up. That made it easy for both people and AI tools to repeat an old assumption, miss a trade-off or present test evidence as a production outcome.
Feature purpose, user and operator needs, included and excluded scope, decision history, trade-offs and known gaps.
Clear source-confidence categories, production evidence versus QA or test evidence, and explicit notes where adoption or business impact was not measured.
Engineering, UX, QA and support handoffs, acceptance criteria, failure states, operational history and the source links needed to verify a claim.
Human review, attribution, uncertainty labels and secret exclusion before context is reused in a PRD, ticket, test plan, research task or handover.
Used across
One supporting workflow: YAML PRDs
Rob introduced the concept. Thuta built the generator. I adopted, operationalised, tested and refined the AI-assisted workflow on live product work. No measured time, quality or defect improvement is claimed. The useful lesson was narrower: generated requirements became safer when the workflow retrieved product, repository, schema and design-system context first, then required a person to review the result.
I use this record for requirements, QA planning, research and handover. It is designed to support onboarding, but I do not claim team-wide onboarding use or adoption. This is evidence of a working product-operations practice, not proof that AI improved delivery speed or defect rates.
Systems I’ve worked with
I have used these systems directly while testing APIs, mapping data, setting up workflows and supporting live products.
Customer records, trading data, segmentation and workflow inputs.
OAuth, verification, rewards and platform-specific API constraints.
Store logic, external catalogue experiments and governed review flows.
API testing, workflow automation and support operations.
Other platform work
Smaller examples that show how I handle identity, quality, client operations and external constraints.
Experience
Product experience across B2B2C platform systems, client integrations and operational delivery.
May 2024 to Present
Product Manager
First and sole Product Manager for a configurable B2B2C loyalty platform, working across client integrations, administrative tooling and member reward experiences. I introduced clearer acceptance criteria, operational visibility and durable product context alongside delivery.
Dec 2021 to Jan 2024
Financial Advisor
Built client communication and decision-making discipline in a regulated environment, including a scenario-based financial planning tool.
Contact