Built on 25 years of thermoforming expertise. Trained on yours.
Built on 25 years of thermoforming expertise.
Trained on yours.
thermoform.ai runs on CognitionWorks' Sprocket AI platform.
The combination gives you an industry knowledge base built from the GTTI library and a private vault where your own documentation lives, searchable and secure.
thermoform.ai runs on CognitionWorks' Sprocket AI platform.
The combination gives you an industry knowledge base built from the GTTI library and a private vault where your own documentation lives, searchable and secure.
How it works
The GTTI Library
The GTTI Library
The GTTI Library
The industry's most comprehensive thermoforming knowledge base.
The industry's most comprehensive thermoforming knowledge base.
Mark Strachan built the Global Thermoform Training Institute over 25 years. Thousands of documents, troubleshooting guides, process tools, resin preparation protocols, tooling guidance, and training materials — all specific to thermoforming. No general-purpose AI model carries this depth. It is the content backbone of thermoform.ai.
Mark Strachan built the Global Thermoform Training Institute over 25 years. Thousands of documents, troubleshooting guides, process tools, resin preparation protocols, tooling guidance, and training materials — all specific to thermoforming. No general-purpose AI model carries this depth. It is the content backbone of thermoform.ai.
Your Private Vault
Your Private Vault
Your plant's knowledge. Yours alone.
Your plant's knowledge. Yours alone.
Upload your own documentation. SOPs, maintenance logs, quality manuals, short videos, machine schematics. Your content is encrypted, isolated, and never shared with other companies on the platform. Every conversation is logged and searchable — turning daily interactions into a continuously growing institutional knowledge asset.
Upload your own documentation. SOPs, maintenance logs, quality manuals, short videos, machine schematics. Your content is encrypted, isolated, and never shared with other companies on the platform. Every conversation is logged and searchable — turning daily interactions into a continuously growing institutional knowledge asset.
How Queries Work
How Queries Work
Ask a question. Get a sourced answer.
Ask a question. Get a sourced answer.
Ask in plain language. The platform searches both the GTTI library and your own documentation simultaneously. Every answer comes with a citation pointing back to the source material. If the answer came from an outdated or superseded document, the trace report makes that visible.
Ask in plain language. The platform searches both the GTTI library and your own documentation simultaneously. Every answer comes with a citation pointing back to the source material. If the answer came from an outdated or superseded document, the trace report makes that visible.
Voice & Video Capture
Knowledge capture that matches how people actually work.
Consumers routinely watch YouTube tutorials to fix household appliances. Manufacturers have been slow to apply the same logic to production equipment. Workers record a short voice or video clip explaining what they did and why. The AI synthesizes the recording into searchable content. No Word docs required. No documentation sprints before someone's last day.
The Sample Query
Your plant's knowledge. Yours alone.
Query: "Can I use PET tools for PP parts?"thermoform.ai: Yes, but you need to compensate for PP's lower shrinkage uniformity. Adjust index and floating knives for trim registration, retune cooling channels for PP's lower thermal diffusivity, verify vent sizing, and check knife length against trim-press tonnage for thicker PP runs. See GTTI sec. 4.2 for full guidance.
Architecture
Built for scale. Precise at retrieval.
3 steps:
- Ingestion: parse your documentation library.
- Semantic indexing: context-aware search that finds "motor" when you search "servo."
- Precision retrieval: the right paragraph, not a summary of the chapter. The platform leverages LLMs from frontier models, wrapped in architecture designed for manufacturing specifics.
Voice & Video Capture
Knowledge capture that matches how people actually work.
Consumers routinely watch YouTube tutorials to fix household appliances. Manufacturers have been slow to apply the same logic to production equipment.
Workers record a short voice or video clip explaining what they did and why. The AI synthesizes the recording into searchable content. No Word docs required. No documentation sprints before someone's last day.
The Sample Query
Your plant's knowledge. Yours alone.
Query: "Can I use PET tools for PP parts?"
thermoform.ai: Yes, but you need to compensate for PP's lower shrinkage uniformity. Adjust index and floating knives for trim registration, retune cooling channels for PP's lower thermal diffusivity, verify vent sizing, and check knife length against trim-press tonnage for thicker PP runs. See GTTI sec. 4.2 for full guidance.
Architecture
Built for scale. Precise at retrieval.
3 steps:
Ingestion: parse your documentation library.
Semantic indexing: context-aware search that finds "motor" when you search "servo."
Precision retrieval: the right paragraph, not a summary of the chapter.
The platform leverages LLMs from frontier models, wrapped in architecture designed for manufacturing specifics.