Immune Super Intelligence
Scientific reasoning · Evidence intelligence · GraphRAG · AI orchestration · Evaluation · Research workflows
The Vision
Making the immune system understandable, connected and investigable.
We are building an evidence-aware scientific intelligence platform that connects immunology's mechanisms, diseases, therapies, data and research into one explorable system.
Move immunology from fragmented information to connected understanding — and from connected understanding to scientific discovery.
Learn. Explore. Investigate. Discover.
ISI + IMMUNE CITY
The world, and the intelligence that navigates it.
AI sits on top of the knowledge and evidence — never in place of it.
Immune City
A connected immunology atlas where users explore cells, molecules, receptors, pathways, mechanisms, diseases, biomarkers, therapies, trials and research.
- Learn Understand biology
- Explore Trace mechanisms
- Investigate Develop hypotheses
02 / THE CAUSAL SPINE
Connect the entire chain.
Not a decorative diagram — a structure you can traverse. Pick a stage.
03 / WHAT MAKES ISI DIFFERENT
Connected, inspectable scientific reasoning.
A connected biological model
Not another collection of articles — a multi-layer model connecting entities, mechanisms, contexts and consequences.
Evidence-aware scientific intelligence
Answers trace back to real sources, distinguishing association from causation and hypothesis from validated finding.
An investigation workbench
Map mechanisms, compare explanations, identify evidence gaps and prepare experiments for human review.
Learning through exploration
Build causal understanding by navigating biology and testing reasoning, not memorising isolated facts.
04 / ROADMAP
Four connected stages.
Strategic ambitions — not claims that every capability exists today.
- Stage 1
Build the scientific foundation
Ontology, knowledge graph, typed relationships, provenance, evidence labels and curated content.
Where we are: In progress — curated graph and evidence lifecycle live; citation coverage still sparse.
- Stage 2
Make the knowledge explorable
Cell atlas, mechanism explorer, disease and therapy connections, causal pathways, learning journeys.
Where we are: Early — graph, mechanisms and Academy foundation track available.
- Stage 3
Add evidence-grounded intelligence
Reliable retrieval, GraphRAG, cited Q&A, evidence comparison, uncertainty-aware synthesis.
Where we are: Prototype — Ask retrieves from the graph with guardrails; not yet fully evidence-grounded.
- Stage 4
Support scientific investigation
Hypothesis generation, multi-source synthesis, dataset analysis, experiment planning — with human review.
Where we are: Ambition — research workbench is an early scaffold.
05 / WHO IT'S FOR
One foundation, many workflows.
- Students and educatorsLearn immunology as a connected system.
- Immunologists and biomedical researchersExplore mechanisms, literature and unresolved questions.
- Translational scientists and drug-discovery teamsConnect mechanisms with biomarkers, targets, therapies and trials.
- Computational biologists and AI researchersCombine structured knowledge, datasets and AI methods.
- Scientific institutions and collaboratorsContribute reviewed knowledge, evidence and workflows.
06 / WHAT ISI IS NOT
Not just another…
- An AI chatbot with an immunology theme.
- A conventional course website with videos and quizzes.
- A graph visualisation with no meaningful scientific provenance.
- A literature search engine that returns papers without connecting their mechanisms.
- A clinical diagnosis or treatment tool without validation, governance and regulatory assessment.
07 / NORTH STAR
From a question to a reviewed hypothesis.
Why might an inflammatory pathway drive disease — and is it a useful target?
- 01
Explore the relevant cells, receptors, signalling pathways and context.
Open → - 02
Trace the proposed causal mechanism through Immune City.
Open → - 03
Inspect the evidence supporting and challenging each relationship.
Open → - 04
Connect the mechanism to disease, biomarkers, therapies and trials.
Open → - 05
Identify missing evidence and competing explanations.
Open → - 06
Formulate a research hypothesis and a possible experiment.
Open → - 07
Return new findings to the research workflow for review.
Open →