In additive manufacturing (AM) in aerospace and defense, documentation overhead is one of the most persistent barriers to getting parts into production. Boeing was sitting on a substantial backlog of technical data packages for casting and 3D printed parts, at a cost its engineering teams could no longer absorb. Authentise, a Philadelphia-headquartered digital workflow management company, worked with Boeing to deploy an AI-assisted documentation tool that Boeing estimates will save approximately $8 million in engineering costs in its first year of deployment.
Simon McCaldin, Open Innovation Lead at Authentise, laid out how that system works at the AMA: Aerospace Space and Defence 2025 event.
Register for AMAA 2026, taking place online on July 9th.
The paperwork problem behind AM adoption
Authentise has operated in the AM sector for twelve years, with staff working in the US and the UK. Its software portfolio spans engineering collaboration, production workflow management, and parts cataloguing, all built around what McCaldin described as “capturing and exploiting the digital thread in manufacturing”.
Bringing a part into production requires translating unstructured engineering intent, covering CAD files, performance specifications, and design rationale, into technical artefacts that multiple stakeholders can act on. McCaldin put a number to that process: 120 to 150 hours per part, even for relatively simple assemblies. “It’s a pain in the backside,” he said, “for anybody who’s had to be involved in anything like this before.”
From collaboration tool to AI-assisted documentation
Threads, Authentise’s engineering collaboration platform, was developed to apply agile methodologies to research and development (R&D) activities. It functions as a structured workspace combining messaging, project management, document repositories, and 3D model annotation. The platform is fully permissioned, making it usable across supply chains and between organisations.
The first AI layer added to Threads is a large language model (LLM) running in the background that analyses what engineers type and suggests how to log it, flagging when a message reads like a decision or action item and prompting the user to record it as such. McCaldin compared it to a helpful version of Microsoft’s Clippy for Engineers. Engineering discussions logged this way become consistently structured, which matters for what the platform does next.
ThreadsBot builds on that structured record, allowing users to query the entire workspace and retrieve synthesised summaries of project status, open issues, and prior decisions. On a Department of Defense (DoD) contract where Authentise served as prime contractor, with around 50 engineers from six organisations contributing, ThreadsBot surfaced three significant project risks that a formal model-based systems engineering (MBSE) approach had missed. McCaldin also highlighted its value for onboarding: new stakeholders can interrogate the workspace directly rather than pulling engineers away from active work.
$8 million in projected savings at Boeing
ThreadsDoc automates the generation of technical data packages (TDPs) from material held within the Threads environment. Users select the relevant threads and data sources, write prompts defining what each section of the output document should cover, and the system assembles a draft report from those inputs.
Authentise developed and tested this with Boeing, which was managing a backlog of TDPs across casting and 3D printed parts. Boeing’s assessment is that the tooling will save approximately $8 million in engineering costs in its first year of deployment. McCaldin was specific about where those savings come from. The tool handles the retrieval and assembly work, pulling together the latest versions of specifications and relevant documents, which engineers find time-consuming but low-value. “What it can do is the legwork of bringing all the documentation together under your guidance,” McCaldin said, “and then output a document, which is simple to review and double check.” Sign-off remains with the engineer.
An industry converging on the same problem
Authentise’s approach targets the gap between where engineering knowledge accumulates, spread across emails, meetings, PLM systems, and legacy drawings, and where it needs to end up: in structured, queryable documentation. ThreadsDoc addresses TDP generation as the highest-friction point in that process, using AI to handle the assembly work so engineers can focus on review and verification.
The documentation challenge extends beyond any single company. ThreadsDoc’s formal launch in November 2024 came at a point when the AM industry was already treating workflow automation as a prerequisite for scaling production, with qualification increasingly identified as the next frontier. At the research level, Fraunhofer IAMHH‘s ADAPT project, AI-Driven Aerospace Part Toolkit, is working to automate finite element model creation to accelerate the certification of complex aerospace components — addressing the same downstream bottleneck from a different angle. The shared aim across both efforts is reducing manual overhead at engineering transition points while keeping human judgment intact at sign-off.
Certification standards will determine how far AI can go
Asked during the Q&A session whether text-to-3D systems could eventually displace design and manufacturing engineers, McCaldin pointed to the certification requirements governing safety-critical production. “You have design authorities who need to sign off on this,” he said, “and it’s going to be quite a long time before we move away from that requirement.”
Authentise’s near-term development priorities reflect that reality. The company is working to connect Threads to external product lifecycle management (PLM) systems so engineers can query those repositories without migrating data into the platform. Work is also underway on extracting design intent from legacy drawings, files that predate digital workflows and may carry annotations representing institutional knowledge with no living custodian. A third workstream, described as in early stages, involves using AI to identify parts within existing catalogues that are candidates for AM conversion.
McCaldin’s closing argument pointed to what he sees as the field’s most important unsolved problem: connecting process simulations and digital twins of manufacturing processes to verified production records. “Tying up process simulations with digital twins of processes with what was actually manufactured, and making sure those are tied up together,” he said. “If we can solve that problem in the next couple of years, then that will make huge inroads for the application of additive in a number of different markets.”
Register for AMAA 2026, taking place online on July 9th.
To stay up to date with the latest 3D printing news, don’t forget to subscribe to the 3D Printing Industry newsletter, follow us on Twitter, or like our page on Facebook.
While you’re here, why not subscribe to our Youtube channel? Featuring discussion, debriefs, video shorts, and webinar replays.
Featured image shows Additive Manufacturing Advantage: Aerospace, Space and Defense. Image via 3D Printing Industry.
