Find Solutions

This video provides a walkthrough of the Find Solutions agent within Eureka, which applies TRIZ methodology to help solve technical problems using cross-domain and current-domain innovations. You'll learn how to structure a prompt around a property you want to change, its context, and any constraints, and how to refine that problem through the agent's follow-up questions before starting the analysis. You'll also see how the resulting mind map breaks solutions down by problem direction, showing the TRIZ principles, patent evidence, and detailed solution information behind each suggestion. By the end of the video, you'll understand how to fine-tune or fuse solutions together and share your findings with your development team.

• From the Eureka home page, click Find Solutions in the center of the page.
  
  This opens the Find Solutions agent.
  
  The agent is designed to help solve technical problems using TRIZ methodology.
• Start with a how-to question.
  
  Specify the property you want to improve and the direction of change.
  
  Example: increase energy density.
  
  Add the context for the problem.
  
  Example: lithium batteries.
  
  Include the constraint or contradiction.
  
  Example: increase energy density without increasing battery volume.
  
  Decide whether to keep cross-domain applications turned on.
  
  Turn it on to get ideas from other industries.
  
  Turn it off if you only want current-domain solutions.
• Click the button on the right to start the process.
  
  The agent will ask clarifying questions about your problem.
  
  Use the plus button to add suggested answers into the response box.
  
  Continue answering until you run out of useful context or want to move on.
• If you want to stop the question flow, click Start Analysis.
  
  This tells the AI that it has enough context and should begin generating solutions.
  
  The agent then creates a problem analysis summary.
  
  Review this summary carefully to make sure it matches your intent.
• The agent extracts the technical parameters.
  
  It shows the desired improvements and the constraints side by side.
  
  It then proposes several solution directions.
  
  You can:
  
  keep all directions,
  
  narrow them down to the most relevant ones,
  
  or add your own direction in the input box.
  
  When ready, click Explore Solutions.
• After a few minutes, the agent generates a mind map.
  
  The left most section shows the technical problem overview.
  
  This is followed by the selected problem directions.
  
  Each direction branches into solution paths informed by TRIZ and supporting evidence.
  
  Click into any section to inspect the underlying reasoning.
• Open the principle section to see which TRIZ principle was used.
  
  TRIZ uses 40 principles to guide problem solving.
  
  The example shown is Local Quality.
  
  The agent explains the principle with a simple analogy and shows patent evidence supporting it.
  
  This helps you understand why the solution direction was chosen.
• Click View Solution Details to open a dedicated page.
  
  Review the high-level summary of the proposed solution.
  
  Check the expected effect and risks.
  
  Read the additional context, including:
  
  technical principles involved,
  
  industry applications,
  
  reliability,
  
  technical maturity,
  
  expected development challenges.
  
  Review supporting references, similar patents, similar literature, and related projects.
• If a solution is especially relevant, click it to fine-tune it for your application.
  
  Use multi-solution fusion to combine multiple promising solutions into one output.
  
  Use the Share button in the top-right to send the results to your team.
  
  This makes it easier to align on the same data before development begins.