01
Start with the part CAD
Bring the geometry, material, production volume, and critical requirements into one engineering brief.
WORKFLOW STUDY / 01
Built with manufacturing engineers.
AI for injection mold design.
We’re developing AI to turn part CAD into mold designs and auxiliary tooling. Our goal: less repetitive engineering work, a shorter path to production.
In development. Defining design pilots with tooling teams.
Names on this page include Antler, Powell St.
Part CAD → Mold designIn development
Core. Cavity. Auxiliary tooling.
Workflowillustration
From part CAD
to mold design.
The workflow
we’re building.
02
Review manufacturability
Identify features that need attention before tooling, from wall thickness to draft and undercuts.
WORKFLOW STUDY / 02
03
Develop the mold and tooling
Explore core and cavity geometry, cooling, ejection, and auxiliary tooling around the part.
WORKFLOW STUDY / 03
04
Validate with engineers
Review the design against the shop’s requirements and compare time and quality with its current process.
WORKFLOW STUDY / 04
Built around real workflows.
We’re defining design pilots with engineers to learn where AI can save the most time.
Built for tooling teams
Your engineering knowledge. Less repetitive work.
Mold design depends on experience, judgment, and the realities of each shop. We’re working with engineers to understand what can be automated and where their expertise matters most.
Built around your shop.
Every team has its own tooling standards, equipment, and ways of working. Our aim is to support existing manufacturers with AI that accounts for those constraints.
Engineers guide the design.
We’re developing a workflow where engineers can review proposals, make decisions, and validate results. Design pilots will help us compare engineering time and quality against the current process.
The engineering we’re exploring
We’re learning from mold designers where time and cost accumulate. These are the questions guiding our research and design pilots.
Engineering question
Which features could create problems during molding?
We’re investigating how geometry checks can surface thick sections and abrupt transitions early, with enough context for engineers to decide what needs to change.
Engineering question
How will this part release from the mold?
We’re exploring how AI can assess draft against the pull direction, material, and surface requirements, and flag areas for an engineer to review.
Engineering question
Where should the mold split around this part?
We’re studying how engineers choose parting lines and handle undercuts, balancing part appearance, tool complexity, and access.
Engineering question
How should cooling fit around the mold geometry?
We’re mapping the constraints behind cooling layouts, including access, available space, and interference with other tooling components.
Engineering question
How can the part be removed without damaging it?
We’re learning how engineers plan ejection around the part’s geometry, surface requirements, and the rest of the mold assembly.
Engineering question
What else does the shop need to make this design work?
We’re looking beyond the core and cavity to the auxiliary tooling and shop-specific constraints that turn a mold concept into a complete engineering project.
Frequently Asked Questions
Luphra is a Dutch, Antler-backed startup based in San Francisco. We’re building AI for injection mold design and auxiliary tooling, working with manufacturing experts to understand their workflows and define design pilots.
We’re developing AI to help tooling teams go from a part’s CAD file to a reviewed mold design. Our focus includes manufacturability analysis, core and cavity geometry, and the supporting tooling. The aim is to reduce repetitive engineering work within existing tooling shops.
It is in development. We’re defining pilot scopes with manufacturing partners and testing where AI could save engineering time. The workflow illustrations on this site show what we’re building toward, not a finished product or measured performance results.
We’re speaking with mold designers, tooling shops, and manufacturers who work on injection-molded parts. We want to understand where design, review, and revisions take the most time, and which problems are worth solving first.
Yes. Engineer review is central to the workflow we’re developing. A proposed design must be checked against the part requirements, the shop’s processes, and the intended production conditions.
Our approach is factory-agnostic: we intend to support existing tooling shops and their engineering workflows. We’re using discovery and pilots to understand the data, software, and manufacturing constraints an integration would need to support.
design pilot
Yes. We’d like to learn how your team works, where time and cost accumulate, and what a useful improvement would look like. That conversation helps us identify whether there is a suitable task to explore together and how we would evaluate it.