Selection and adaptation, not invention
A catalogue of the solution shapes we work in — recurring end-to-end patterns for solving a class of problem — organised by the data-flow shape of the system underneath. The asset behind our approach.
How it’s used
New clients are classified against known shapes rather than diagnosed from zero, then matched to a starting solution shape. The match happens in the synthesis phase of our methodology — it’s the hinge that lets us prototype rapidly without sacrificing fit.
Shapes are organised by what the system actually does with data — publish it, capture it, transform it, decide on it, adapt to it, or act on it autonomously. That structure is the point: picking the right shape is picking the right architecture.
This is a separate asset from the skills library. Solution shapes are not built from skills, and skills do not roll up into solution shapes.
33 shapes
Pitch & Concept Validation Gallery →
“We need to show investors what we're building before a single line of product code is written”
Interactive Editorial & Data Story →
“We have a powerful dataset but publishing it as a PDF report means nobody reads it”
Field & Offline Data Collector →
“Our staff work in areas with no mobile signal and data just disappears”
Internal Lightweight Tracker / Replace-a-Spreadsheet →
“We're running everything on a shared spreadsheet and it keeps breaking”
Structured Intake Form & Workflow →
“People email us their requests and they get lost in the inbox”
Multi-Tenant Operational Capture SaaS →
“We want to build a SaaS product for our industry but don't know where to start”
Real-Time Collaborative Workspace →
“Two people edited the document at the same time and now we have two conflicting versions”
Order & Transaction Capture →
“Orders come in by email and phone and we lose track of which ones have been fulfilled”
Automated Recurring Report Generator →
“Our analyst spends two days every month just pulling the same report together”
KPI Dashboard & Data Refresh Pipeline →
“Someone pulls the numbers into a spreadsheet every morning and it's already out of date by lunch”
Alerting & Exception Detection Engine →
“We find out something went wrong hours after it happened”
Data Integration, Sync & Workflow Orchestration →
“We export a CSV from one system and manually import it into another every week”
Data Cleaning & Canonicalisation Pipeline →
“No one trusts the data because the same entity appears under ten different names”
Prioritisation & Ranking Engine →
“We have hundreds of sites and no principled way to decide which ones to tackle first”
Guided Protocol / Decision-Tree Assistant →
“Our staff are skipping steps or going off-script in the field”
Calculator / Estimator Tool →
“Our advisors are doing this calculation on a spreadsheet and emailing it to clients”
Scoring & Classification Engine →
“Our team is manually reviewing every application and we can't keep up with volume”
Matching & Ranking Engine →
“We have a large pool of options and no quick way to surface the best fit for a given request”
Knowledge Q&A Assistant (RAG / Document Intelligence) →
“Staff waste hours searching through policy documents to find a specific answer”
Next-Best-Action Recommender (Stateless) →
“Our frontline staff don't know what to do next and they default to the same script for everyone”
Document Generation & Templated Synthesis →
“Our team spends hours manually drafting contracts or proposals that are 80% the same every time”
Adaptive Learning & Practice System →
“Our learners all move at the same pace even though their gaps are completely different”
Spaced-Repetition & Retention Engine →
“Our users pass the initial training but forget most of it within a few weeks”
Personalised Recommendation & Feed Engine →
“We have thousands of items in our catalogue but users only ever find the same ten”
Behaviour-Change & Coaching Loop →
“Users start strong but disengage after two or three weeks and don't come back”
Dynamic Routing & Matching Engine →
“We're still manually assigning learners to tutors based on spreadsheets and gut feel”
Personalised Content & Curriculum Sequencing →
“New users go through the same onboarding regardless of their role and most of it isn't relevant to them”
Single-Task Action Bot →
“We keep assigning someone to do the same task over and over — it's just clicking buttons”
Research & Synthesis Agent →
“Our analysts spend two days pulling together a brief that should take twenty minutes”
Conversational Action Agent →
“Our chat widget answers questions but can't actually do anything — customers still have to call us”
Multi-Step Process Automation Agent →
“We have a process with ten steps across four systems and someone has to shepherd every case through it”
Event-Triggered Autonomous Action Agent →
“We get the alert but by the time someone acts on it the damage is already done”
Multi-Agent Orchestrated System →
“The problem is too big and too varied for one model to handle well end-to-end”