Keiji AI LogoKeiji AI

Research Camp 2

Oncology data science camp

A four-week program in which oncology data scientists use TrialMind on real multi-omics and clinical data, producing demos, publications, and product feedback with our team.

May 11 – Jun 1, 2026Mondays 9–10 AM or 9–10 PM EST4 weeks

Example analysis

TrialMind reproducing a published PLOS Medicine analysis in non-small cell lung cancer (NSCLC), an example of the projects in this camp.

Camp outputs

Participant use cases, a publication track, and a structured evaluation of TrialMind.

Joint publications

Co-authored papers on the methods and findings produced in the camp.

Participant demos

Recorded TrialMind demos built from participants' own use cases.

Community stories

Posts from participants sharing what they built and learned.

A user survey

Structured feedback on workflow fit, time saved, and trust in outputs.

Product improvements

A prioritized list of TrialMind enhancements from real-use friction.

Schedule

Four weeks with weekly office hours, from preparation to a final demo day, with a deliverable each week.

  1. Week 0

    Preparation

    • Finalize datasets (e.g., TCGA) and task definitions
    • Approve participants and provision TrialMind access
    • Prepare tutorial materials and baseline examples
  2. Week 1

    Kick-off and onboarding

    • Kick-off: capabilities, workflow demo, datasets
    • Hands-on tutorial session

    Deliverable: A project plan per participant

  3. Week 2

    Exploration

    • Run cases in TrialMind on chosen tasks
    • Office hours: progress, Q&A, debugging

    Deliverable: Preliminary results and an issues list

  4. Week 3

    Development

    • Iterate with deeper analyses
    • Office hours: progress, Q&A, debugging

    Deliverable: A draft demo and early insights

  5. Week 4

    Demo Day

    • Final demos and survey collection
    • Recap and joint write-up planning

    Deliverable: A recorded demo and program summary

Project scope

Projects are scoped so each participant completes a demo and a write-up in four weeks.

Datasets

  • Multi-omics cancer datasets such as The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO)
  • Clinical data layered in where available
  • Public, well-documented sources to reduce friction

Tasks

  • Aligned to TrialMind strengths
  • Biomarker discovery, subtyping, survival prediction, risk stratification
  • Scoped to be achievable in three weeks

Evaluation

  • Biological interpretability of results
  • Clinical relevance for downstream decisions

Deliverable format

  • A demo combining slides with the live system
  • A report template built for publication reuse

Register interest

Places are limited. We follow up with dataset options, task scoping, and onboarding.