Frequently Asked Questions
Thermoform.ai is an AI knowledge platform built exclusively for thermoforming converters. These are the questions we hear most often from process engineers, plant managers, and operations leaders considering the platform.
What is thermoform.ai?
thermoform.ai is a domain-specific AI knowledge platform built exclusively for thermoforming converters. It gives operators, process engineers, and plant managers instant, cited answers to process questions — 24 hours a day, 7 days a week — drawing from a library of over 25 years of structured thermoforming expertise developed by Mark Strachan of the Global Thermoform Training Institute (GTTI). Unlike general-purpose AI tools, thermoform.ai is trained on thousands of thermoforming-specific documents: troubleshooting guides, process specifications, material behavior data, tooling references, and industry training materials. It runs on the CognitionWorks Sprocket AI platform and can also be trained on your own company documentation.
What is the difference between Sprocket and thermoform.ai?
Sprocket is the underlying AI platform developed by CognitionWorks. thermoform.ai is the domain-specific content library — thousands of documents, presentations, guides, and calculations covering thermoforming processes — delivered through Sprocket. Think of Sprocket as the engine and thermoform.ai as the specialized knowledge pack. The pilot program gives thermoformers access to the thermoform.ai content library via Sprocket, with the option to upload your own documentation to preview the full platform capability.
How is thermoform.ai different from ChatGPT or other general AI tools?
thermoform.ai answers thermoforming-specific questions with precision that general-purpose AI tools cannot match. A tool like ChatGPT can respond to broad manufacturing questions, but it cannot troubleshoot film gauge variation on a specific forming application or advise on cavity layout validation for a PET tray run. The core differences are: Source specificity: thermoform.ai draws from GTTI's 25+ years of structured thermoforming content — tables, charts, troubleshooting guides, and process tools built for the industry. Citation traceability: Every answer points back to source material, so operators can verify the response against original documentation. Your own documentation: The platform can ingest your SOPs, machine manuals, maintenance logs, and process recipes, so answers reflect your plant's specific processes — not generic guidance. No hallucination risk from generic training: General AI models generate plausible-sounding answers from broad internet training data. thermoform.ai constrains responses to vetted, domain-specific source material.
Who is Mark Strachan and what is GTTI?
Mark Strachan is the founder of the Global Thermoform Training Institute (GTTI), with over 25 years of hands-on thermoforming training and process expertise. The GTTI knowledge base — thousands of documents, guides, calculations, and training materials — forms the foundation of the thermoform.ai content library. It is the most comprehensive structured thermoforming knowledge library available in AI form.
How does the platform answer questions?
You ask questions in plain English — the same way you would text a colleague. thermoform.ai searches the GTTI knowledge base and your own uploaded documentation, then returns a precise answer with a citation pointing to the source material. For example: asking "Can I use PET tools for PP parts?" returns a specific answer covering shrinkage compensation, cooling channel adjustments, vent sizing, and trim press considerations — with a reference to the relevant section in the source guide.
Does the platform support multiple languages?
Yes. thermoform.ai supports instant translation into dozens of languages, accessible with a single click. This is particularly valuable in production environments where operators and engineers may not share a first language. A question asked in Spanish, French, or Portuguese returns an answer from the same verified knowledge base — reducing troubleshooting delays caused by language barriers.
Can I upload my own company documents?
Yes. The platform supports upload of your own manuals, SOPs, maintenance logs, process recipes, and video recordings. Once uploaded, the AI can draw on your documentation alongside the thermoform.ai content library to return answers specific to your equipment and processes. During the pilot, users receive one document, manual, or video upload per month as a preview of this integration capability.
What about voice and video capture?
Operators and engineers can record short voice or video clips explaining how they diagnosed a problem or completed a repair. These recordings are ingested by the platform and become searchable content — turning informal knowledge into a queryable internal resource. A 30-second explanation from an experienced technician becomes as findable as any written procedure.
Does the platform integrate with ERP or production systems?
Full ERP, MES, SQL, and network drive integrations are available through the complete Sprocket platform. The pilot program provides a limited preview of this capability, focused on the thermoform.ai content library and document uploads. Integrations with systems such as IQMS, shared drives, and production databases are scoped during onboarding for post-pilot engagements.
What is the onboarding and training process?
Setup requires no local software installation. Access is cloud-based and works on any device — desktop, tablet, or phone — with a browser and login credentials. The onboarding process includes: Kick-off call — CognitionWorks walks through platform access, upload limits, and how to get the most from the thermoform.ai content library during the pilot. Immediate access — No pre-loading of your own data is required. The full thermoform.ai GTTI knowledge base is available from day one. Document upload guidance — The team helps you identify which of your own documents would be most valuable to add during the three-month pilot. Ongoing support — CognitionWorks collects feedback throughout the pilot to identify gaps and improvements in the knowledge base. User limits do not exist. Companies may want internal policies around access management, but there is no per-seat charge or restriction on how many people can access the platform.
What is included in the pilot program?
The three-month pilot provides: Full access to the thermoform.ai content library (the complete GTTI knowledge base) Limited Sprocket platform functionality, including one document, manual, or video upload per month Cloud-based access for unlimited users across up to two sites A kick-off session and ongoing support from the CognitionWorks team No ERP or production system integrations are included in the pilot. These are available in post-pilot commercial engagements.
How much does the pilot cost?
The pilot is priced at USD $4,500 for three months, with unlimited users and a maximum of two sites. The three-month commitment gives users sufficient time to explore the content library, integrate the platform into daily workflows, and help identify gaps or opportunities for improvement. There is no deadline for signing up — the clock starts when you enroll — though pilot slots may be capped.
What happens after the pilot?
Post-pilot pricing depends on the level of integration required and the number of sites enrolled. There is a one-time integration fee that varies based on the number of connected systems (ERP platforms, network drives, SQL databases, IQMS, and similar), plus a monthly recurring fee covering both Sprocket and thermoform.ai access.
Is there a deadline to sign up?
There is no fixed deadline. The pilot clock starts from your enrollment date. CognitionWorks may cap the number of concurrent pilots, so early enrollment is recommended.
How is my company's data protected?
thermoform.ai uses enterprise-grade security architecture. Your uploaded data is stored in a dedicated, isolated vault — separated from every other company on the platform — and encrypted both at rest and in transit using AES encryption and TLS 1.3. The security architecture has been audited and approved by Coca-Cola, Baxter International, and BDO. Key protections include: Isolated data vaults: Each company's data is stored separately. No data is shared between tenants. Zero Trust architecture: Server-to-server access requires authentication at every step, limiting exposure within the network. Role-based access control (RBAC): Client administrators control which users can access which documents and which content the AI is permitted to reference. SOC2-aligned posture: All infrastructure providers are SOC2 current and compliant
Where is data stored?
All AI memory and client data is stored in a secure AWS data center in the United States (Virginia). The platform runs on Microsoft Azure infrastructure with Cloudflare Web Application Firewall (WAF) protection at the network perimeter.
Are the AI models trained on my data?
No. The Large Language Models used within the Sprocket platform are not trained on customer data. Instead, Sprocket uses a neural memory system with security metadata — your documents are stored in your encrypted vault and retrieved at query time, constrained to your user's access permissions. Your data never becomes part of a shared training set.
What subprocessors does the platform use?
CognitionWorks uses the following subprocessors, all SOC2 and ISO27001 certified, with data located in the USA: Amazon AWS — infrastructure and storage OpenAI — LLM processing (no data retention) Google Cloud Vertex — LLM processing (no data retention) Anthropic — LLM processing (no data retention) Cloudflare — network security and WAF (no data retention)
What happens to my data after the pilot ends?
All data — including uploaded documents and chat history — is securely wiped at the end of the pilot period unless you request a secure export. If you continue as a full subscriber, your data carries forward into your ongoing account.
How do users log in? Is there a separate password to manage?
No separate password is required. SprocketAI uses Microsoft OAuth2 security tokens via your existing corporate Microsoft Entra (formerly Azure Active Directory) identity manager. Users authenticate with their existing corporate credentials. This gives your IT team a centralized point of control and ensures that deactivated accounts lose access immediately.
How do I start the pilot?
Contact the thermoform.ai team directly to discuss your pilot. Email: cc@thermoform.ai