Every example here is a finished study: a decision an engineer owns, the method that settles it, and a part that ships with it. Run one as it stands and you have a real result in minutes.
Then order it your way. Say what is different about your part, your material or what you are optimising, and it adapts the use case. The method and the guardrails stay put.
Python 3.12, 3.13 or 3.14 · Windows / macOS / Linux · the first one needs no account at all
Each example names the decision it answers, what it costs to run, and where to point it at your own part.
Open a finished result on the real part - surface or draggable solid cut, time slider, field toggles - and pull the numbers behind it as plain arrays. A reference result and its geometry ship with the example, so it runs before you have an account.
The whole loop in one script: mesh locally, upload the geometry, choose a material, run AI filling, read the peak cavity pressure. Swap in your STL and it is your part. The result comes back viewer-ready.
Gate count, then placement, then process, then validation. Every candidate scored for filling pressure, fill evenness and weld lines - and the reason behind each cut written to the project's decision log as it is made.
Which parameters matter, where to centre them, how wide the window may be. One config file drives a wide sweep over three variables, screening, adaptive refinement at the feasibility boundary and a set of numerical confirmation decks - then emits the spec and the audit trail behind it.
One call over the STL: wall thickness, flow path, flow length and air traps, each finding carrying the place on the part it came from. The limits are yours to set, and the versioned report names the checks that did not run rather than leaving them out.
Bring it to a community session and walk us through it. The people in the room are working on the same problems, and what you learned the hard way saves someone else a week.
If you have never worked with agentic tooling, you are not alone. Most of our customers had not either when they started with the toolkit. We wrote down what you need to get going: which agent to run, three steps for your first session with the prompts to paste, and what usually goes wrong.
One subscription covers up to 3 people · monthly, cancel any time.