In custom manufacturing, a new engineer can read every drawing and still not know how to cost the job. The specification is on the page. What it means for routing, tolerances, yield and margin is not. That understanding takes years to build, and most of it is learned beside a senior colleague.
A companion piece covers how to keep senior engineers' reasoning as the work is done. This one covers what that record makes possible: bringing new engineers to independent, reliable work much sooner.
Why expertise in evaluation is slow to transfer
The philosopher Michael Polanyi observed that people know more than they can put into words. Experienced engineers are a clear case. Ask why they chose a routing and the answer often starts with "it depends", followed by a list of considerations they weigh without thinking. That knowledge is tacit. It does not survive being written into a manual, because the manual captures rules, while the expertise lies in knowing which rule applies, and when to break it.
Nonaka and Takeuchi's research on knowledge creation explains how organisations normally move tacit knowledge between people: mostly through shared experience, working side by side, before anything is written down. In manufacturing, that means a junior engineer shadowing a senior one, job after job. It works. It is also slow; it depends on the senior engineer's time, and it stops when that person is unavailable.
What apprenticeship gets right
Collins, Brown and Newman's model of cognitive apprenticeship describes why learning beside an expert works. The expert first models the task, then coaches the learner through it, provides scaffolding where the learner is unsure, and gradually fades that support as competence grows. Ericsson's research on deliberate practice adds that expertise grows fastest with frequent, specific feedback on real tasks.
The problem is not the method. It is that each new engineer can only learn from the jobs and the seniors they happen to meet.
Structured precedent as scaffolding
When evaluations and costings run through a system that preserves the reasoning, every past job becomes a worked example. A new engineer facing an unfamiliar enquiry no longer starts from a blank cost sheet. They see how comparable parts were routed, which tolerances drove cost, where past estimates were revised, and why.
The system can play the parts of the apprenticeship model at scale:
Modelling: showing how the firm handled similar jobs, with the reasoning attached.
Coaching: asking the specific questions an experienced engineer would ask at each decision.
Scaffolding: pre-filling what is routine and flagging what needs judgement.
Fading: asking fewer leading questions as an engineer's evaluations hold up on review.
The new engineer still makes the decisions. They simply make them with the firm's accumulated experience in front of them, rather than whatever one mentor remembers.
Senior engineers become reviewers and coaches
This changes how senior engineers spend their time. Instead of walking a junior colleague through routine evaluations, senior engineers review, correct and explain the hard cases. Each correction becomes part of the record, so the next engineer learns from it too. Mentoring stops being one-to-one and becomes something the whole team benefits from.
What to track
Manufacturers adopting this approach can measure whether it works: the time until a new engineer completes evaluations without heavy revision, how much rework senior review still finds, and what share of enquiries new engineers can take on independently. Those measures show whether the precedent is doing its job.
How Mitochondria's agentic AI systems support engineering teams
The approach. The quotes, technical offers, and costing evaluation units (units 61, 62, and 63) are configured to each manufacturer's processes, rates, and templates. Decision memory, part of the shared foundation, keeps the reasoning behind every evaluation in the manufacturer's own environment.
What it enables. New engineers work with the firm's own precedent, guided question by question, from their first week. Senior engineers review and coach on the cases that need them. The business can see how its evaluation capability grows.
The impact. New engineers contribute real work sooner. Senior engineers spend their time where their judgement matters. And the firm's know-how stays with the firm as people join, move and retire.
Read more about manufacturing.
Sources
Polanyi, M. (1966). The Tacit Dimension.
Nonaka, I. and Takeuchi, H. (1995). The Knowledge-Creating Company.
Collins, A., Brown, J. S. and Newman, S. E. (1989). Cognitive apprenticeship: teaching the crafts of reading, writing and mathematics.
Ericsson, K. A., Krampe, R. T. and Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance.
—
Mitochondria builds agentic AI for evidence-led core operations. Research-led and model-agnostic. ISO/IEC 27001:2022 certified, with controls aligned to the GDPR, UK GDPR, India's DPDP Act and Australia's Privacy Act. Amsterdam and Pune. Request a working session →