Track 01 · Rejuve.bio — Biomedical research assistance

Build an Omega research assistant that finds and explains studies about medicines or lifestyle changes affecting a biological pathway.

Who you are building for

Rejuve Biotech is building an AI research platform to speed up discovery of therapies for aging and age-related disease. Its approach unifies genetic, clinical and preclinical data into a knowledge graph, then uses neural-symbolic AI to generate testable hypotheses. Your users are researchers who need answers they can trace and check.

The problem

Research findings are scattered across papers. Studies may investigate different populations, measure different outcomes or reach different conclusions. Researchers need an assistant that preserves these differences and explains what the evidence actually supports.

Your build

An Omega agent that takes a pathway and one focused question — for example "Which medicines or lifestyle changes have been studied for their effects on insulin signaling in humans, and what did the studies find?"

For trainees: insulin signaling is the communication process through which insulin — a hormone — tells cells how to take up, use and store sugar. Think of a pathway as a chain of biological messages.

Omega should discover relevant papers and return:

• the interventions studied, with source links and supporting quotations
• who participated, what was tested and what it was compared with
• what researchers measured and what changed
• whether the biological explanation was tested or remains a hypothesis
• conflicting findings, reported harms and missing information
• a conclusion explaining uncertainty and what evidence would change it

Lower blood sugar alone does not prove which part of insulin signaling changed.

Keep the scope narrow: choose medicines or lifestyle changes, focus on human studies, and inspect up to 10 papers.

What Omega does here

Agent loop	Plans searches, calls tools, inspects results and decides whether to investigate further or finish.
Memory and reasoning	Saves checked sources, reuses findings across sessions and compares evidence across study contexts.
Tools	Public literature search, source retrieval and structured evidence storage/export.

Use Omega's native orchestration and persistent memory. PLN and numerical scoring are not required.

Demo script (3 minutes)

1. Ask the focused research question.
2. Show the source-linked findings, explanation, uncertainty and tool execution.
3. Start a new session, ask a follow-up and demonstrate memory reuse.
4. Show conflicting findings and an example where evidence is insufficient.

Showcase checklist

• Every substantive research claim links to supporting evidence
• Scientific terminology is accurate, with brief explanations where needed
• Measured findings and proposed explanations are distinguished
• Conflicting and missing evidence are handled explicitly
• Omega's decision loop, tool execution and memory reuse are demonstrated

Boundaries

Use public, synthetic or explicitly authorised data only. No personal health data. This is research support, not medical advice, and the interface must say so.

Stretch ideas

Extract numerical effects and uncertainty; identify papers reporting the same study; export the evidence trail.

Best for

Builders who like evidence, retrieval and agent workflows.