What Hestus is building
Hestus takes a different approach from prompt-driven CAD generation. It describes its product as 'Autocomplete for CAD': software that watches the context of the designer's current work and predicts a likely next modeling action. The company emphasizes one-keystroke suggestions inside the CAD environment rather than moving the user into a separate prompt-first interface.
Where it sits in the engineering workflow
This resembles the productivity pattern that made code completion useful to software developers. A mechanical engineer still owns the feature tree and decides design intent, while the assistant tries to remove repetitive interaction around obvious next steps. That narrower scope can be attractive in professional CAD because it reduces the risk of an agent inventing an entire geometry from an underspecified prompt.
Why mechanical engineers should care
Mechanical CAD contains many micro-actions that are individually small but collectively expensive: selecting constraints, closing profiles, adding repeated geometry and navigating commands. If prediction is reliable, saving a few seconds per action can matter across thousands of interactions. The company currently advertises productivity gains on its site, while also noting that some demonstrated beta features are available only to selected users.
What it does not remove from engineering
The usefulness of autocomplete depends on precision. A suggestion that is wrong often enough can interrupt concentration and create more checking than it saves. It also has to respect enterprise requirements around local data, model confidentiality and supported CAD versions. Users should evaluate it on representative production models rather than demonstration sketches alone.
Questions to ask before adopting it
- What percentage of suggestions are accepted on real production work?
- Which CAD systems and modeling operations are supported today?
- Can suggestions be disabled or constrained for sensitive/critical workflows?
- Where is model context processed and how is proprietary geometry protected?
DIVE-LD engineering view
The most useful way to evaluate Hestus Autocomplete for CAD is to put it inside a controlled engineering release process. Ask whether the output preserves design intent, whether assumptions are visible, whether another engineer can edit or audit the result, and what verification remains before manufacturing. AI can compress repetitive work and widen the design space, but responsibility for requirements, safety, standards, tolerances and final release still belongs to the engineering organization.
Sources and further reading
Public product information reviewed on 11 September 2026. Company statements about performance or capability are vendor claims unless independently verified.