Applied AI · 2024
TurningPoint
An AI-powered decision support platform that organises complex information and presents structured insights without replacing human judgment.
01 · The problem
Why this was
worth building.
Important decisions can involve more information and competing factors than a person can comfortably process at once. The project explored how structured workflows and AI-assisted analysis could make that information easier to examine.
02 · What I wanted to achieve
Collect the context needed to understand a decision
Present AI-generated insights in a structured, readable way
Support the user’s judgment instead of trying to replace it
03 · System architecture
How the pieces
work together.
04 · Technical decisions
Guide the user before analysing
An interactive workflow collects relevant inputs before requesting AI analysis, giving the backend more useful context than an open-ended prompt alone.
Connect the interface to AI services
The React frontend communicates with FastAPI services through REST APIs and turns returned analysis into components users can read and explore.
Make the output easier to act on
Structured recommendations and interactive visualisations help users compare information without losing control of the final decision.
05 · Project targets
These are design targets and project measures, not claims about production usage.
The hardest part
The challenge was not only receiving an intelligent response from the backend. The interface also had to organise that response so users could understand the reasoning and decide what to do with it.
The choice I made
A guided workflow asks for more effort than a blank chat box, but it gives the analysis better context and makes the result easier to follow.
What I learned
TurningPoint introduced me to integrating AI into a real application. I learned that useful AI depends as much on clear presentation and user trust as it does on generating an intelligent output.