PRODUCT MANAGEMENT · PAYMENTS · REGULATED SAAS · APPLIED AI
Complex problems.
Clear product decisions.
I'm Rhys Dowsall, a Product Manager and Product Owner with seven years' experience turning customer, commercial and regulatory complexity into practical outcomes across financial services and SaaS.
Targeting Product Manager and Product Owner roles in regulated fintech and SaaS · also open to Senior Business Analyst opportunities · Belfast / UK remote
SELECTED IMPACT
SELECTED WORK
Decisions, constraints and outcomes.
Not a catalogue of job descriptions — a few examples of the problems I've worked through and the decisions behind them.
01 · PAYMENTS & SAAS · PLOTBOX
Scaling digital payments—without scaling the risk
As Product Owner for the Finance stream, I worked across payment links, card and ACH workflows, bulk processing, settlement and reconciliation as PlotBox Pay moved into a higher-volume phase.
The problem
Payment capability had to support more volume and more complex customer workflows while preserving financial controls, clear status visibility and a safe release path.
Product decisions
I shaped a card-first bulk-payments MVP, separated ACH as a follow-on, defined confirmation, permission and reporting safeguards, and used feature flags and controlled customer validation.
Delivery constraints
The work crossed customer operations, finance, Engineering and external payment services. Security gaps and reconciliation needs meant release readiness mattered as much as feature scope.
Programme outcome
Monthly PlotBox Pay throughput grew from approximately $230k to $1.5m over roughly six months. That is programme-level context—not a claim that one feature caused the full increase.
02 · TECHNICAL PRODUCT & BUSINESS ANALYSIS · NETSUITE
Turning a nine-step finance process into an auditable integration
An enterprise customer needed to transfer PlotBox financial data into NetSuite. The existing process involved nine manual steps, nominal-code mismatches, repeated adjustments and developer involvement whenever mappings changed.
My role
As Product Owner for the Finance stream, I led discovery and business analysis with the customer's Finance and IT teams, translated the existing process into requirements, coordinated backlog and testing activity, and maintained the decision and risk record.
Key decisions
We defined a one-way journal-export MVP, customer-managed Chart of Accounts mappings, aggregation rules by transaction type and payment method, and approval and date-sensitive configuration controls.
Safe delivery
The API-based workflow was validated against a NetSuite sandbox using representative batches. CSV export remained available as a testing, reconciliation and audit safeguard.
Validated outcome
More than 480 test journal batches were processed in approximately 20 minutes. The project reached end-to-end testing readiness; production benefits had not yet been measured.
Constraints: legacy finance logic, multi-system mappings, date-dependent approval rules, API constraints and the need to preserve a safe fallback.
03 · DIGITAL BANKING · AVANT MONEY
Stronger authentication without losing sight of customers
As Digital Product Manager, I owned digital banking journeys affected by PSD2 Strong Customer Authentication while balancing regulatory delivery, customer friction and longer-term product quality.
The decision
I mapped the affected journeys, coordinated an interim OTP solution and helped shape biometric authentication as the stronger long-term experience.
What we measured
The change was delivered to the fixed regulatory deadline. We tracked OTP conversion and drop-off rather than treating compliance as a binary delivery exercise.
Wider digital ownership
During my broader ownership of online banking and the mobile app, the app rating rose from 3.2 to 4.5 over six months. This is product-level context rather than a claim that one change caused the full increase.
Consent at scale
I led the OneTrust rollout across seven public websites. Regression and remediation reduced more than 20 unauthorised cookies to zero, while evidence-led copy testing was followed by approximately 85% opt-in.
04 · AI & PRODUCT DISCOVERY
AI Product Discovery Assistant
I built a working review prototype that turns unstructured customer feedback into evidence-linked findings and proposed opportunities, with human review kept explicit.
Try the public review workflow using synthetic feedback, evidence-linked findings and explicit human accept/edit/reject decisions.
PRODUCT LAB · PRODUCT JUDGEMENT
A decision, not a demo.
A short, illustrative scenario about releasing AI into a regulated customer journey.
There is no score and no trick answer. The point is to make the trade-offs visible—and show the questions I would ask before committing to a release.
ILLUSTRATIVE SCENARIO · REGULATED CUSTOMER SUPPORT
An AI assistant is ready to answer customer account questions.
The pilot data looks promising, but some incorrect answers concern payment dates, balances and arrears.
94%overall accuracy
+17%customer satisfaction
2.1%high-risk answer errors
−22%assisted contacts
What would you do?
Choose a release approach, or skip straight to the reasoning.
MY PRODUCT APPROACH
Your call:
Pilot with controls—and make the risk measurable.
The aggregate result is encouraging, but accuracy alone cannot justify a broad release when a small number of failures could cause financial harm.
- Block autonomous answers about balances, arrears and payment commitments.
- Route uncertain and high-impact cases to a person.
- Measure errors by severity and customer impact, not only volume.
- Define audit, rollback and stopping thresholds before expanding.
Questions before release
Where do errors cluster? Who can be harmed? Can every answer be traced? How quickly can the feature be stopped?
The goal is not to avoid risk entirely. It is to preserve the customer value while containing, observing and learning from the risk.
HOW I WORK
Evidence before theatre.
I tend to gravitate towards messy product problems: payments that don't reconcile, processes with too many manual steps, requirements that aren't as simple as they first appear, and questions about where AI genuinely helps.
I make the failure mode visible, separate evidence from assumptions and turn ambiguity into decisions a team can deliver. In regulated work, controls, exceptions and release criteria are part of the product—not paperwork added at the end.
RELEVANT PRODUCT & DIGITAL EXPERIENCE
Independent product workAug–Oct 2025Discovery and requirements for contractor software
PlotBoxJul 2024–Jul 2025Product Owner — Finance
Avant MoneyFeb 2019–Jul 2024Digital Product Manager
TeleperformanceFeb 2017–Nov 2018Service Delivery — Vodafone Ireland GovCare
First Derivatives / KXJun 2016–Feb 2017Business Development
LexisNexis2007–2016Legal Research & Training
LET'S TALK
Interesting product problem?
I'm always happy to talk product, payments, digital transformation, applied AI and the awkward problems in between.