Use case
Auto-triage RFQs. Enforce your shop’s rules.
Stop chasing infeasible drawings.
Werk24 reads technical drawings and compares critical specs to your machine and process constraints. Catch infeasible requests early, propose feasible alternatives, and focus engineers on winnable work.
The bottleneck in feasibility checks
Triage of incoming drawings is still manual in many shops. Engineers read PDFs, decode tolerances, check threads and materials, and then call customers to renegotiate. It is slow, inconsistent, and expensive — and the worst RFQs steal time from the winnable ones.
What it checks — your rules, codified
- Tolerances — min/relative widths, IT grades, general tolerances; unit-aware.
- Threads — type (metric/UTS/Whitworth), pitch, class, min nominal Ø, depth.
- Radii & chamfers — minimum radius/width; honors general tolerances.
- GD&T — min tolerance per characteristic (position, flatness, perpendicularity…).
- Materials — whitelist/blacklist by standard or types (e.g., “no aluminium” or “only polymers”).
- Surface finish — min. allowed roughness (e.g., Ra/Rz) or N-grade.
Scope rules per part family, customer, or machine group. Combine constraints to match how you actually build parts.
How it works
Define your own decision rules and workflow.
Document your feasibility catalogue You
Capture your feasibility checks in a human-readable spec. We help ensure the rules stay consistent and conflict-free.
Design your workflow You
Decide whether RFQs forward to a dedicated Werk24 inbox or integrate directly with your ERP, CPQ, or ticketing stack.
Quality assurance Werk24
Provide 30 representative drawings with expected outcomes. We monitor accuracy continuously so new releases never break your flow.
Go live
Roll out the automated triage and monitor exceptions in real time.
Outcomes
Say no faster to the wrong work — and yes faster to the right work.
Minutes to Triage
Auto-check drawings and return pass/fail with explanations in seconds.
Consistent Rule Enforcement
Codify know-how: the same rules apply across teams, shifts, and sites.
Protect capacity
Filter out non-winnable RFQs and prioritize profitable ones.
Traceable rationale
Every decision is explainable through item-level checks and an audit trail.
Metrics that matter
Track whether feasibility checks are faster, clearer, and more effective.
- Triage speed: Time-to-decision (P50/P90) per RFQ.
- Auto-pass rate: % of drawings that pass without human review.
- Exception load: % flagged; average violations per drawing.
- Turnaround: Time from customer RFQ → first response with suggestions.
- Resolution rate: % of suggestions accepted by customers.
FAQ
How do we model our shop’s reality?
Begin with a core set of constraints that reflect your machines and QA rules. Expand gradually by adding part families or customer-specific policies. Scoping is flexible—apply rules per customer, machine group, or globally.
Which standards are supported?
We normalize common thread systems (Metric, UTS, Whitworth), tolerance formats including general tolerances, and key GD&T symbols. Units are auto-detected, and each extraction comes with a confidence score.
Can I get annotated PDFs back?
Yes. Alongside the JSON API output, you can request annotated PDFs that highlight each violation and propose feasible alternatives. Attach them directly to your RFQ workflow.
What if confidence is low?
All low-confidence items are automatically flagged for human review. You control the thresholds and decide how items should be routed — for example to estimators, QA, or customer teams.