Product Owner & Developer

Designed,
owned, shipped.

Three software products where I held the full product owner and developer role simultaneously. Backlog, architecture, user decisions and deployment — all one job.

Products 3 shipped
Domain Fintech · AI · iOS
Stack Full-stack + Native
Role PO + Developer
01
Financial Intelligence · 2024–2025

InsideStockData

Full-stack platform for institutional SEC filing intelligence and money flow tracking

Live SEC / EDGAR Node.js PostgreSQL Firebase Auth Stripe RSS Pipelines
Live
InsideStockData platform dashboard

A financial intelligence platform ingesting SEC filings (10-Ks, 10-Qs and 13Fs) via asynchronous RSS pipelines and surfacing institutional money flow data in real time. Every design and development decision was mine: data contracts, refresh cadences, leaderboard design, monetisation architecture. The platform includes community mechanics and Stripe-integrated subscription billing.


3 Filing types ingested
(10-K · 10-Q · 13F)
Real-time Institutional money
flow visualisation
Async RSS pipeline
architecture
Stripe Subscription
monetisation

Key Features

InsideStockData pipeline
Feature 01

Async SEC Filing Pipeline

Asynchronous RSS ingestion of SEC EDGAR filings with structured data contracts and configurable refresh cadences per filing type.

InsideStockData institutional flow visualisation
Feature 02

Institutional Flow Visualisation

Real-time visualisation of institutional money flow derived from 13F filings, surfacing position changes, new entries and liquidations across a watchlist.

InsideStockData community and monetisation
Feature 03

Community & Monetisation

User leaderboard mechanics and Stripe-integrated subscription billing. Tiered access model designed to convert engaged free users into paying subscribers.

02
AI Governance · 2025–present

AIXER

EU AI Act compliance platform with multi-factor risk scoring and enterprise integrations

In Development TypeScript React Auth0 PostgreSQL LeanIX ServiceNow
In Development
Aixer platform dashboard

An AI governance platform built for EU AI Act compliance, where I acted as product owner and sole developer. I defined the compliance scope, designed the data model and built the system. The core is a multi-factor risk scoring engine across five independent dimensions, backed by a granular RBAC model, bi-directional integrations with LeanIX and ServiceNow, and an append-only digitally signed audit log. The architecture is pre-designed for PostgreSQL to MongoDB migration as data volume scales.


5 Risk scoring
dimensions
RBAC Granular role-based
access control
2 Enterprise integrations
(LeanIX · ServiceNow)
Signed Append-only
audit log

Key Features

Aixer Risk Matrix
Feature 01

Multi-Factor Risk Scoring

Five independent scoring dimensions mapped to EU AI Act risk categories. Each dimension is configurable per organisation context, with weighted aggregation into a single compliance risk score.

Aixer Data Compliance
Feature 02

Bi-directional Enterprise Integrations

Two-way sync with LeanIX for enterprise architecture context and ServiceNow for ITSM workflow triggers. AI system records in AIXER reflect live data from both platforms without manual entry.

Aixer Audit Trail
Feature 03

Digitally Signed Audit Log

Append-only event log covering assessments, status changes, access decisions and integration events. Each entry is digitally signed to meet regulatory tamper-evidence requirements under the EU AI Act.

03
Native iOS · 2025–present

TourTour

Proximity-based landmark discovery with AI-generated audio guides and hybrid caching

Final Deployment Swift MapKit Auth0 iOS LLM CoreLocation
Final Deployment
Swift Tourism App screenshot

A native iOS app where I built both the product definition and the development. As product owner I set the core problem to solve: how to deliver LLM-generated audio guides in real time without the latency and cost of per-step API calls. As developer I built the answer by developing a hybrid caching architecture that pre-generates and stages content for likely nearby landmarks using MapKit and CoreLocation proximity data, delivering a fluid consumer experience without compromising on personalisation.


MapKit Native proximity
detection
Hybrid On-device + cloud
caching architecture
AI Audio LLM-generated
landmark guides
Real-time Location-triggered
content delivery

Key Features

Tour Tour Geofetching
Feature 01

Proximity Detection via MapKit

CoreLocation-powered geofencing triggers landmark detection as users move. MapKit renders nearby points of interest with context-aware overlays that update in real time without manual interaction.

TourTour Audio Guide
Feature 02

AI-Generated Audio Guides

LLM-generated audio narration adapts to the specific landmark, time of day and user context. Each guide is generated once, cached locally and served instantly on subsequent visits.

TourTour Caching Architecture
Feature 03

Hybrid Caching Architecture

Pre-generation and local staging of content for landmarks within a predictive proximity radius. Solves the LLM latency problem for a consumer app without removing the personalisation that makes AI guides worthwhile.

Get in touch

Want to walk through any of this in more detail?

I am happy to share more on architecture decisions, product choices, or anything else across these three projects.