From Undocumented Codebase to AI-Powered WealthTech Platform
A cloud-native OCIO platform combining microservices architecture, a full backend migration, and a live AI investment co-pilot -delivered under investor and revenue timelines.
Outcome Snapshot
5
user journeys delivered and live
4
knowledge domains covered by co-pilot
2 days
defect resolution, down from 20
<20s
AI co-pilot response time, every interaction
Client
A UK-based, VC-backed WealthTech startup building an AI-powered OCIO (Outsourced CIO) platform for family offices, trustees, and UHNW/HNW clients. The platform was at an early stage when Systango was engaged -built by a prior vendor, largely undocumented, and facing dual deadlines: a production MVP to begin generating revenue and an investor-facing prototype for an upcoming fundraise.
Challenge
-
No handover documentation: Systango inherited the platform with no architecture, business-flow, or developer setup documentation from the outgoing vendor.
-
Infrastructure security gaps: public-facing services without a WAF and unencrypted RDS instances needed remediation immediately.
-
Two parallel workstreams: a production MVP with five core user journeys had to ship alongside a separate investor-facing clickable prototype.
-
An AI co-pilot under a hard latency bar: Stella had to stay responsive under 20 seconds across multiple knowledge domains.
Objectives
-
Take over the existing platform and establish full technical ownership
-
Run a full technical and infrastructure assessment to identify and remediate gaps
-
Execute a backend tech-stack migration from Java/Spring Boot to Node.js
-
Build and deliver five core user journeys within the first sprints
-
Design an investor-facing clickable prototype without pulling focus from the MVP
-
Integrate Stella, an AI investment co-pilot, across four knowledge domains
-
Establish production-grade cloud architecture on AWS and remediate security gaps
What We Built
- Full technical and infrastructure assessment: identifying and remediating security gaps including private subnet migration, WAF implementation, and RDS encryption.
- Complete backend tech-stack migration: from Java/Spring Boot to Node.js across all core services.
- Core platform modules: Model Portfolio, My Info, My Office, Product Screener, Macro Library, IPS Monitoring, and Spaces.
- An investor-facing clickable prototype: designed and delivered on a parallel track to the production MVP.
- Stella, an AI investment co-pilot: integrated across Strategy, Economic Forecast, Market Watch, and Spaces, using OpenAI and Refinitiv data feeds on AWS Lambda.
- A microservices architecture: on AWS ECS Fargate with a Vue.js SPA frontend and Cognito-based authentication.
- AI-assisted engineering practices: used across design, development, and QA, with every AI-generated output reviewed by a developer before release.
- Governance and compliance controls: role-based access control, audit trails, and release approval gates, built in from day one.
Impact
Deployed
AI co-pilot Stella live, responding in under 20 seconds
Delivered
five core user journeys, enabling revenue generation
Accelerated
AI-assisted development cut defect resolution time by 90%
Governed
AI outputs & infrastructure through RBAC, audit trails, and security remediation
Build Facts
Platform
Web App (Vue.js SPA)
AI Co-Pilot, Cloud Backend (AWS)
Investor-Facing Prototype
Focus
AI Co-Pilot Integration
Backend Tech-Stack Migration
Investment Portfolio Management
Infrastructure Security
Investor Prototype Delivery
Let’s talk, no strings attached.
