Redko Labs editorial notes · Updated October 12, 2026
Week one: map a real workflow
Choose one potential user group. Write down how they handle the problem today, what it costs in time or resources, who pays, and who can approve a change. Ask for recent examples rather than asking whether they like your idea.
Week two: identify the ownership questions
Before sharing confidential work or promising a company exclusive rights, identify contributors, employment or student agreements, grants, patents, datasets, software licenses, and university policies. Ask your technology transfer office which rights could be licensed and what approvals are needed. Ownership cannot be inferred from who wrote a prototype.
Week three: test one assumption
Build the smallest experiment that can change your mind. For an AI tool, define an evaluation set and a baseline. For scientific software, identify the calculation, experiment, or decision that should improve. Record failures and limits as carefully as successes.
Week four: choose a next step
Compare the technical evidence with customer evidence. Continue if there is a specific problem and a plausible route to adoption; narrow the scope if the technology works but the buyer is unclear; stop an idea when the next experiment is unlikely to answer a useful question. A negative result can be a good outcome.
Common questions
Does joining an incubator transfer my university research?
No application should be treated as a transfer. Any required license or assignment needs to be identified and documented with the actual rights holder.
Can I explore this while studying?
You can apply while studying. Participation must fit your university, employment, funding, and work-authorization obligations.
Further reading
Program details are proposals until confirmed in an offer. These notes are general information, not legal, investment, medical, or immigration advice.