Intro call
A first conversation about your business and your ask.
30 minutes
Services
From first conversation to running system.
Our services follow one workflow: Discover, Plan, Deliver, Support. Each phase stands on its own, so you can start where you are and continue as far as you need.
How we work here
A clear path through the work.
Start with an initial call. Each phase can stand on its own.
- 01
Discover
- 02
Plan
- 03
Deliver
- 04
Support
At a glance
Choose the service you need.
Discover
Intro call
30 minutes
Discovery session
2–3 hour session · written report to follow
Plan
Define
Decide what to build
1–3 weeks
Assess
Review what you have
1–3 weeks
Deliver
Build
New software, built from scratch
Feature: 1–3 weeks · MVP: 3–6 weeks · Products: 6–12+ weeks
Evolve
Improve an existing system
Sprint: 1–4 weeks · Program: 1–3 months
Support
Advisory
Senior guidance over time
Monthly engagement
Run
Care after launch
Monthly engagement
01 / Discover
Start with the right questions.
Discovery session
A working session that turns your needs into requirements, a draft architecture and an iteration plan. You receive a Discovery Report with goals, a requirements outline, a rough plan and the recommended next step.
2–3 hour session · written report to follow
In detail
For example
- “I have an idea for a new product”
- “We need a trading screen or OMS: where do we start?”
- “Our ops team runs on Excel”
- “We want AI in research but don't know what's safe”
02 / Plan
Decide what to build.
Define
Plan something new. A Blueprint sets out what to build, how, in what order and at what cost. Choose a Feature Blueprint for one capability, or a Product Blueprint for a standard or large platform.
1–3 weeks
Feature Blueprint · 1 week
One capability, scoped and specified.
Product Blueprint · 2 weeks
A product or platform of standard size.
Large Product Blueprint · 3 weeks
A large product or platform, such as a full trading system.
In detail
For example
- A new internal platform: priorities, architecture, data design and roadmap
- AI platform design: RAG, agents, orchestration and monitoring
- Post-trade architecture: client accounts, statements, transfers, analytics and audit
- Trading system architecture: execution, orders, events and storage
- Client statements or an audit-log feature
- Exchange integration
- Position transfers or an approval workflow
- Event-driven architecture and audit-trail design
Assess
Review an existing system, engineering health or AI opportunities. The Diagnostic report gives you findings, a risk view and a prioritised roadmap.
1–3 weeks
In detail
For example
- AI readiness: Data, security, governance, guardrails and the opportunities worth pursuing.
- AI engineering assessment: How the team works with AI tools and agents, and what to roll out.
- Rescue assessment: What's broken, what to keep and what to rebuild — including an AI-built prototype that needs a real architecture.
- Performance and scalability review: Where a system slows down or falls over under load, and what to fix first.
- Engineering health: Codebase, testing, delivery pipeline, monitoring and team productivity.
03 / Deliver
Make it real, then make it better.
Build
New products and capabilities, from one missing feature or a focused MVP through to a larger platform. Built from an approved Blueprint, in milestones — by us, or handed to your team to build.
Feature: 1–3 weeks · MVP: 3–6 weeks · Products: 6–12+ weeks
Feature build · 1–3 weeks
The one missing capability, designed and built.
MVP · 3–6 weeks
One core workflow, a few users, essential integrations.
Product build · 6–12 weeks
Several workflows, roles and permissions, real-time data and reporting.
Large product build · 12+ weeks
Multiple integrations, event-driven, high reliability, AI agents with oversight.
Every build includes hypercare after launch — defect fixing and launch questions in business hours, sized to the build.
In detail
For example
- A brand-new product built from the ground up, with AI at its core
- A trader dashboard with positions, P&L and alerts
- An AI research assistant over internal data
- An ops platform: portal, admin, reporting and integrations
- A research platform: news, AI summaries and collaboration
- A portfolio management platform: client reporting, performance and analytics
- An internal admin platform: CRM, compliance, finance and ops
- An end-to-end trading platform with exchange connectivity
- Multi-vendor real-time market data
Evolve
Improve an existing platform. An Evolution Sprint tackles a focused batch of features, UX, technical debt or performance work. An Evolution Program brings a series of sprints together around a roadmap.
Sprint: 1–4 weeks · Program: 1–3 months
Evolution Sprint · 1–4 weeks
A fixed batch of features, UX, technical debt or performance work.
Evolution Program · 1–3 months
A planned series of sprints against a roadmap.
Hypercare included after launch.
In detail
For example
- Product expansion: a big new capability on an existing platform
- Taking an AI-built prototype to production
- A UX refresh of a trader workflow
- Fixing a buggy, fragile system and its underlying architecture problems
- Performance or scaling fixes when a system slows under load
- Fixing data and reconciliation breaks: wrong numbers, P&L mismatches
- Cutting runaway cloud or AI model costs
- Consolidating scattered dashboards and admin tools
- Front-end modernisation or a market data feed revamp
- Taking over from a previous vendor and stabilising
04 / Support
Keep the system in good hands.
Advisory
Ongoing senior guidance at three levels, from a senior sounding board to hands-on technical leadership.
Monthly engagement
Advisor · 8 hours a month
Founders and small teams who need a senior sounding board.
Technical Partner · 10–20 hours a month
Pre-seed and early teams actively building, with regular decisions to make.
Fractional CTO · 20–40 hours a month
Companies that need hands-on technical leadership shaping the roadmap and the team.
Every level includes a weekly check-in, support in business hours, architecture and decision reviews, hiring interviews and code review of critical components — and more as needed.
In detail
For example
- Fractional technical leadership: roadmap, hiring, architecture decisions and code reviews
- Investor and board technical support
- Bridging after a key-person loss: stabilise, document and cover until a hire
Run
Monitoring, support and small improvements once the included period after a Build or Evolve ends. Three levels of reserved monthly capacity: Light, Standard and Plus.
Monthly engagement
Available for systems we built or evolved, as an extension of the support already included.
In detail
For example
- Monitoring and alerting
- Model and prompt upgrades for AI features
- Exchange connectivity maintenance
- Fixes and small improvements
Client paths
Ways the services come together.
The paths are optional. They show how different needs use the same services, and how the work continues once a system is running.
Founder path
Founders without an in-house team
Discovery session → Define → Build → Run + Advisory
Growth path
Companies with existing software they want to extend, improve or modernize
Assess → Evolve → Run or Advisory
Feature path
Teams missing one capability
Workshops or Discovery session → Define → Build → Run or Advisory
Flagship platform path
Companies building a large product or platform
Discovery session → Define → Build → Run + Advisory
AI productivity path
Teams that want to work faster with AI
Workshops or Discovery session → Assess → Build → Run
For most projects, we advise a Blueprint comes first. At its approval checkpoint, you can proceed with the Build, reprioritise to fit your budget or request a revised estimate if material new work has emerged.
Our focus
What we don't do.
Marketing websites · WordPress or Shopify builds · Native mobile apps · SEO · Generic cloud migration · Standalone security or DevOps consulting · Staff augmentation · Open-ended bug fixing · Undefined scope · Coding by the hour
Not sure where to start?
An initial call about your business and what you need.