Prompt for refined company search

Chatgpt pro prompt

You are doing a fresh DeepResearch pass to identify pilot companies for RapidDraft.

This is NOT a generic company list task. The goal is to find a very small number of extremely well-matched companies that fit a sharply refined ICP based on our previous research and mistakes.

You must optimize for fit, not volume.

If a company from earlier lists appears again, that is acceptable ONLY if it still passes the stricter filters below. If it fails the new filters, explicitly reject it and explain why.

Context: what RapidDraft actually is

RapidDraft should be treated primarily as a mechanical design review and collaboration platform, closer in spirit to CoLab than to a pure DFM tool.

Its strongest near-term value is:

comparing revisions and telling teams what changed
checking drawings/models for common release mistakes
capturing review comments, issues, and decisions traceably across revisions
helping design, manufacturing, and quality collaborate around CAD and drawings
fitting into NX / SolidWorks workflows, optionally with Teamcenter / PDM / PLM

Do NOT frame RapidDraft mainly as:

a generic AI design tool
a pure DFM checker
a cheap replacement for sophisticated automation stacks
a quoting tool
a full CAD authoring replacement

The best-fit companies are those where the review process is painful, fragmented, and cross-functional.

Core strategic insight from the previous iteration

Earlier lists were too broad and included too many manufacturing-service companies, CNC shops, molders, and other firms that are better fits for a narrow DFM or drawing-analysis pitch than for the original RapidDraft vision.

The better target is a design-owning mechanical product company or machine builder where:

the company owns product definition
internal engineering creates and releases CAD models and drawings
manufacturing, quality, and often suppliers must review and comment on designs
revision churn and release pressure are real
decisions are still scattered across email, meetings, PDFs, screenshots, spreadsheets, or ad hoc workflows
CAD/PLM maturity exists, but there is no strong dedicated review/collaboration layer already in place

Important nuance:
Do NOT search for the least mature companies.
PDM / PLM / Teamcenter / structured revision control are not negatives.
In fact, some maturity is desirable.
The real gap is: they have design/release complexity, but they do not appear to have solved design review collaboration, drawing-checking workflow, or review traceability with a dedicated tool.

Primary company-level ICP

Find companies that match most of the following:

Design-owning mechanical OEM or product company
Mandatory:
they design their own products, systems, or machines
they release their own drawings / BOMs / CAD data
they have an in-house mechanical engineering function

Good examples:

Sondermaschinenbau / special-purpose machinery
industrial machinery OEMs
packaging machinery
factory automation hardware
intralogistics / material handling equipment
pumps / valves / compressors / fluid systems
industrial HVAC / refrigeration equipment
industrial devices / enclosures / thermal hardware
test & measurement / lab equipment
medical devices with strong mechanical content
off-highway / heavy equipment mechanical suppliers
selected automotive Tier-2 / Tier-3 mechanical component suppliers
Strong collaboration and review need
Must show signs of:
multi-stakeholder review across mechanical design, manufacturing engineering, quality, and optionally suppliers
complex assemblies or variant-rich products
repeated design changes / ECOs / revision churn
documentation-heavy release workflows
manufacturing feedback loops that matter
Mechanical-heavy products
Prefer companies whose products contain meaningful amounts of:
machined parts
sheet metal
weldments
assemblies
sometimes molded parts / housings

Important:
Injection-molded parts are acceptable only when part of a broader design-owning product company.
Pure injection molders are not the target.
Pure CNC service companies are not the target.

CAD / PLM signal
Prefer companies with public signs of:
Siemens NX or SolidWorks
Teamcenter, SolidWorks PDM, or another PLM/PDM backbone
CAD admin / mechanical design / drawing / revision-control job postings
engineering workflows that imply structured release processes

Important:
Using NX or SolidWorks is a positive.
Using Teamcenter or PDM is a positive.
This does NOT disqualify them.

Pilot-friendly size and speed
Ideal:
roughly 50–500 employees
or up to ~1000 if still engineering-led and pilot-friendly
enough engineering complexity to feel pain
small enough that a Head of Mechanical Design / Engineering Manager can still move fast

Avoid:

very large enterprises unless a local division clearly operates autonomously
tiny shops with no structured engineering process
Bonus signals
Strong positive but not mandatory:
international business
English-speaking website or hiring material
multi-site engineering or supplier network
exports / global customers
industries where traceability and review discipline matter

Person-level ICP: who matters inside the account

For each company, identify the most likely internal champion / buyer roles.

Primary likely buyer / economic sponsor:

Head of Mechanical Design
Engineering Manager
Leiter Konstruktion
Leiter Entwicklung

Strong secondary sponsor:

Manufacturing Engineering Lead
Industrial Engineering Lead
Fertigungsplanung / Arbeitsvorbereitung
Manufacturing Technology / Industrialization

Strong tertiary sponsor:

Quality Lead
Quality Manager
QA / compliance leader

Possible day-to-day champions:

Lead Mechanical Engineer
Senior Mechanical Designer
CAD / PDM admin
NPI lead

Important interpretation:
The designer feels the pain, but the person who can often say yes is usually Engineering Manager, Manufacturing Engineering, or Quality.

Pain signals to look for

Look for companies where the likely real pain is:

“What changed between revisions?”
“What must be rechecked before release?”
drawing errors / release mistakes / NCRs / rework
review comments getting lost across email, PDFs, meetings, or screenshots
supplier clarification loops
repeated rediscovery of the same issues
weak traceability of review decisions
manufacturing or quality feedback not being carried cleanly into the next revision

This is the key fit signal.
Do not just look for “they manufacture mechanical parts.”
Look for evidence that they review and release complex designs.

Hard exclusion criteria

Exclude companies if there is public evidence that they already use, market, partner with, or prominently signal any of the following categories:

A) Dedicated DFM / manufacturability tools
Examples:

DFMPro
aPriori
Autodesk Moldflow
explicit DFM software or manufacturability platforms
companies publicly advertising DFM consulting / DFM software workflows on the site

B) Dedicated drawing automation / design automation / configurator stack
Examples:

DriveWorks
CAD configurators
knowledge-based engineering
automated drawing generation platforms
rule-based design automation systems
strong “engineering automation” / “configure-to-order automation” signals

C) Dedicated design review / collaboration layer already solved
Examples:

CoLab or direct equivalents
explicit design review platforms
clearly solved CAD review collaboration tooling

D) “Too advanced” public signals
Exclude if the company publicly looks like a showcase customer, reseller, implementation partner, or highly advanced lighthouse for the same category of tooling you are trying to sell.

E) Wrong business model
Exclude or heavily deprioritize:

pure CNC job shops
pure contract manufacturers
pure sheet metal fabrication shops
pure injection molding companies
tooling-only or mold-only service providers
companies that mainly manufacture from customer drawings and do not own product definition

Important nuance on exclusion:
Base CAD + PLM is not a negative.
Advanced automation / DFM / review-platform signals are the negative.

Public competitor / exclusion search instructions

For every candidate, actively search for negative evidence, not just positive fit.

Run checks like:

company name + DriveWorks
company name + DFMPro
company name + aPriori
company name + Moldflow
company name + CoLab
company name + “design automation”
company name + configurator
company name + “knowledge-based engineering”
company name + “drawing automation”
company name + “automated drawing”
company name + manufacturability
company name + DFM
site:companydomain plus the same terms

If clear public evidence exists, reject the company.

Priority sectors to search first

Search these sectors first because they are the closest fit to the refined RapidDraft vision:

Tier 1 priority

Sondermaschinenbau / special-purpose machinery
industrial machinery OEMs
packaging machinery
factory automation hardware
intralogistics / material handling systems
pumps / valves / compressors / fluid systems
industrial HVAC / refrigeration equipment
industrial electronics hardware / enclosures / chassis / thermal hardware
test & measurement / lab equipment
medical devices with strong mechanical content

Tier 2 priority

off-highway / heavy equipment mechanical suppliers
selected automotive Tier-2 / Tier-3 suppliers focused on mechanical components and assemblies

Search terms to use in German and English:

Sondermaschinenbau
Maschinenbau
Anlagenbau
Verpackungsmaschinen
Automatisierungstechnik
Fördertechnik
Pumpenhersteller
Armaturen
Gerätebau
Prüftechnik
Labortechnik
Medizintechnik Geräte
Kältetechnik
Wärmeübertrager
Kompressoren
Greifer
EOAT
Montageanlagen
conveyor systems
machine builder
special purpose machinery
packaging machinery
fluid handling equipment
lab equipment manufacturer
industrial enclosure manufacturer
mechanical OEM

Geography

Primary geography:

Germany first
especially Bavaria and Baden-Württemberg
DACH next
broader Europe only if fit is excellent

Bonus:

companies with English-speaking presence
companies with international customers / suppliers
companies reasonably reachable from Munich are a plus, but fit matters more than distance

Research method

Use official company websites first, then careers pages, job postings, LinkedIn company pages, press releases, directories, and reputable secondary sources.

For each company, verify all of the following as far as possible:

do they own product design?
are they a product company or machine builder, not just a service manufacturer?
do they likely have cross-functional review pain?
is there NX / SolidWorks / PDM / Teamcenter signal?
is there NO strong public evidence of advanced DFM / review / drawing automation tooling?
are they pilot-friendly in size and decision speed?

Do not pad the list with weak fits.
If you only find 8 strong companies, return 8.
Quality over quantity.

Output required

Return 3 sections.

Section 1: Qualified targets
Return only 10–20 companies maximum.
For each company include:

Company name
Website
Country + city
Employee count estimate
What they make, specifically
Why they fit RapidDraft, specifically tied to:
design ownership
collaboration/review need
revision / drawing / release pain
manufacturing-quality-supplier interaction
CAD / PLM signals found publicly
Negative-tool check result:
explicitly state whether you found any DFM / review / drawing-automation signal
if none found, say “no clear public signal found”
Best target roles / likely buyer titles
Likely pain angle for outreach in one sentence
Confidence score:
High
Medium
Low
Source links / citations

Section 2: Borderline candidates
Return a short list of near-misses that almost fit but have one problem.
For each, explain exactly why they are borderline.
Examples:

too manufacturing-service-heavy
CAD stack unclear
too large / too bureaucratic
possible automation signal but not conclusive

Section 3: Rejected / exclude list
This is important.
List companies from prior-style research or newly found companies that should NOT be approached.
For each rejected company, state the disqualifier clearly:

public DFM signal
public design automation signal
already too advanced
pure manufacturer / job shop / molder
wrong ICP for review-platform wedge

Scoring framework

Score each qualified company on a 1–5 scale for:

Design ownership
Cross-functional review pain
CAD / PLM fit
Absence of competitor / automation signal
Pilotability / likely speed
International / English friendliness

Then rank companies by total weighted fit.

Use these weights:

Design ownership = 25%
Cross-functional review pain = 25%
Absence of competitor / automation signal = 20%
CAD / PLM fit = 15%
Pilotability / speed = 10%
International / English friendliness = 5%

Final interpretation rule

Prefer companies that are mature enough to have revision control and release pain, but not so mature that they already have a dedicated collaboration / DFM / drawing-automation stack in place.

Do NOT optimize for “cheaper replacement of an existing tool.”
Do NOT optimize for generic manufacturers.
Do optimize for companies where RapidDraft would feel like the missing review layer around existing CAD / PLM workflows.

The ideal answer should make it obvious why each selected company would plausibly respond:
“This sounds like the missing piece in how we review and release mechanical designs.”