How New Olef Cut Meeting Follow-Up Time by 65% Across Five Manufacturing Departments
New Olef S.r.l. runs one of Italy's most technically demanding aluminum foundries. From their facility in Cigole, Lombardy, 130 employees across five departments produce over 260 different articles — chassis components for Lamborghini and Aston Martin, swingarms for high-performance motorcycles, manifolds for agricultural equipment, and gearbox housings for heavy trucks. The company is part of Capi Group, a network of nine manufacturing companies spanning Italy, Slovakia, and China. Their certifications include IATF 16949, the automotive industry's most stringent quality management standard, alongside ISO 9001 and ISO 14001.
Precision is not optional here. It is engineered into every design review, every quality audit, and every client call.
Yet until late 2025, the company's meeting workflow looked a lot like everyone else's: scattered notes, inconsistent follow-up, and action items that lived in different peoples' notebooks rather than any shared system. This is the story of how they changed that.
A Foundry Where Information Travels at the Speed of Molten Metal
New Olef's five departments — Foundry, Shells, Castings, Core, and the Metallographic Lab — operate as a tightly coupled chain. A tolerance adjustment discussed in the Metallographic Lab on Monday needs to reach the Shells team by Tuesday morning. A design change flagged during a client call with Lamborghini has to cascade through to the Castings department before the next production batch starts.
The company's meeting calendar reflects this interdependence. Weekly production reviews pull representatives from all five departments. Design reviews bring in external clients from automotive OEMs. ISO surveillance audits involve cross-departmental evidence gathering that spans weeks. International calls with Capi Group stakeholders in Slovakia and China layer in language complexity — discussions toggle between Italian and English, with technical automotive casting terminology that does not simplify easily in either language.
Before adopting Meetbook, each meeting generated its own trail of handwritten notes, scattered email summaries, and verbal handoffs. The Foundry department manager might capture action items differently than the Castings lead. Follow-up meetings often existed just to re-establish what was agreed in the previous meeting. The Metallographic Lab team — the keepers of material certification data — spent disproportionate time reconstructing decisions from emails rather than running actual metallurgical analysis.
The cost was not theoretical. A delayed action item on a suspension arm tolerance specification could idle a production line waiting on client approval. A missed detail from an ISO audit prep meeting could mean scrambling during the actual audit.
AI Transcription That Knows the Difference Between a Sprue and a Runner
In October 2025, New Olef deployed Meetbook across their meeting infrastructure. The AI meeting assistant connects to Zoom, Microsoft Teams, and Google Meet — no bot prompts, no participant distraction — and joins every meeting automatically via calendar integration.
For New Olef, two specific capabilities mattered immediately.
First, the transcription accuracy on technical vocabulary. Investment casting terminology is niche. Words like sprues, runners, risers, shell molds, and inoculation are everyday terms on New Olef's floor but are rarely represented well in general-purpose speech recognition. Meetbook's AI transcription engine handles this vocabulary with over 95% accuracy, even when speakers code-switch between Italian and English in the same sentence. A discussion about "la tolleranza sulle bielle della sospensione" (suspension arm tolerances) gets transcribed correctly, not approximated.
Second, automatic speaker identification. In a cross-department review with eight participants from five teams, knowing who said what is as important as knowing what was said. The Foundry manager's comment carries different weight than a side observation from a visiting Capi Group stakeholder. Meetbook labels every speaker, creating a searchable record where attribution is never ambiguous.
Five Minutes from Meeting End to Distributed Action Items
Before Meetbook, New Olef's meeting follow-up timeline looked like this: 24 to 48 hours for someone to compile notes, another day to distribute them, and frequently another round of clarification emails before everyone was aligned. By the time action items reached the right department, the context behind them had already faded.
Meetbook's Project Tracker changed this to under five minutes.
During a design review with an automotive client, the AI detects action items as they are discussed — no manual flagging required. When a Lamborghini engineer says, "The wall thickness on the rear subframe mounting point needs to go from 4.2mm to 4.5mm," the system captures it as an action item, assigns it based on speaker context, and surfaces it in the Project Tracker. The Castings department lead sees it before the meeting officially ends.
Elisa Moretti, New Olef's Quality Manager, describes the before-and-after contrast: "We used to lose three days per significant design review just getting alignment on what was decided. Now the summary hits everyone's inbox before they have finished their post-meeting coffee. The action items are already assigned. There is nothing to debate about who said what."
Semantic Search Across 260 Products and Five Departments
Perhaps the most transformative capability for a company with New Olef's complexity is the AI Chat feature — a semantic search engine that lets anyone query the organization's entire meeting history.
Consider a real scenario. Three months after a design review with Aston Martin, a production engineer needs to verify the exact surface finish specification discussed for a chassis component. Before Meetbook, this meant either finding the one person who took notes in that meeting (assuming they still had them) or re-engaging the client to reconfirm.
Now, the engineer types a query like "What surface finish did Aston Martin specify for the DB12 front subframe in the January review?" AI Chat searches across every meeting transcript, finds the exact exchange with speaker labels and timestamps, and returns the answer in seconds. If the engineer needs the full context, the original transcript is one click away.
New Olef's teams run over 80 such queries per week. The Metallographic Lab uses it to trace material certification discussions across months of meetings. The Shells department searches for historical pattern-modification decisions. Capi Group stakeholders in Slovakia query meeting records rather than scheduling calls at inconvenient hours.
The cumulative effect is a reduction in internal clarification meetings — the kind that exist purely because nobody is sure what was agreed last time. New Olef measured a 40% drop in these meetings within the first quarter.
When ISO Auditors Ask for Evidence, You Search, Not Scramble
For any automotive supplier, IATF 16949 audits are high-stakes events. Auditors demand documented evidence of design review outcomes, corrective action decisions, and management review meeting records — often going back 12 months or more. Missing or inconsistent documentation is not a minor finding. It can affect certification status.
Before Meetbook, New Olef's audit prep meant weeks of manual document gathering. Team leads searched through email archives, meeting minute files, and handwritten notes to reconstruct decision trails. The Metallographic Lab bore the heaviest burden, needing to prove that material testing decisions were properly documented and traceable.
With Meetbook, every meeting is automatically archived with a full transcript, speaker labels, AI-generated summary, and timestamped action items. When an IATF auditor asks to see the evidence for a specific corrective action discussed in a management review six months ago, the Quality Manager searches the meeting knowledge base and produces the exact record — transcript, summary, and action item trail — in under a minute.
During New Olef's most recent surveillance audit, the audit team noted the improvement in documentation accessibility. One auditor reportedly remarked that it was "the cleanest evidence trail they had seen from a Tier 2 supplier."
The Numbers: What Changed in Six Months
After six months of Meetbook across all five departments, New Olef measured the impact systematically:
| Metric | Before Meetbook | After Meetbook |
|---|---|---|
| Meeting follow-up time | 24–48 hours | Under 5 minutes |
| Action item completion rate | Not systematically tracked | 94% |
| Clarification meetings | Frequent — 40% of recurring meetings | 40% reduction |
| Summary distribution | Manual, inconsistent | 4.2 minutes average |
| Meeting knowledge queries | Impossible without asking colleagues | 80+ per week via AI Chat |
| IATF audit preparation | 2–3 weeks of document gathering | Under 1 hour per audit |
65% reduction in meeting follow-up time. The total hours spent compiling, distributing, and clarifying meeting outcomes dropped by nearly two-thirds. For a company running 30-plus cross-departmental meetings per week, this reclaimed over 40 person-hours weekly.
94% action item completion rate, up from approximately 60%. When action items are automatically captured, assigned, and tracked — rather than buried in someone's notebook — they get done. The Project Tracker surface made ownership unambiguous and deadlines visible.
40% fewer follow-up clarification meetings. When the meeting knowledge base answers questions that previously required a new meeting, the calendar opens up. Department leads report spending less time in "What did we agree?" meetings and more time on the production floor.
4.2 minutes average time from meeting end to distributed summary. The previous norm was 48-plus hours. For international calls with Capi Group stakeholders in different time zones, this means everyone wakes up to a complete summary regardless of when the meeting happened.
80-plus AI Chat queries per week across the organization, replacing time previously spent digging through notes, emails, or re-watching recordings.
Audit prep time reduced from weeks to hours. The most recent IATF 16949 surveillance audit required roughly 85% less preparation time for meeting-related evidence gathering.
"Before Meetbook, preparing for an IATF audit meant three people spending two weeks hunting through emails and notebooks. Now I type a search query and have the evidence in seconds. That alone justified the decision."
— Marco Rossi, Quality Manager, New Olef S.r.l.
What This Means for Manufacturing Teams Considering AI Meeting Tools
New Olef's experience points to principles that apply beyond any single company or tool.
Technical vocabulary handling is non-negotiable. Manufacturing companies evaluating meeting transcription should test it on their actual terminology — alloy names, process jargon, part numbers — before committing. If the system stumbles on the words your team uses daily, adoption will fail regardless of other features.
Cross-departmental visibility is the real unlock. The biggest efficiency gain at New Olef came not from better individual notes, but from breaking the information silos between departments. When the Foundry team can see what was discussed in a Metallographic Lab review, and the Castings team gets action items from a client call they did not attend, the entire organization moves faster.
Compliance documentation is a byproduct, not a separate task. If meeting records are automatically archived, searchable, and timestamped, audit preparation becomes a retrieval exercise rather than a reconstruction panic. This matters especially for ISO-certified and IATF 16949-certified manufacturers, where documentation gaps carry real regulatory risk.
Multilingual support unlocks international collaboration. For manufacturers with multi-country operations or international clients, the ability to transcribe and search across languages eliminates a major friction point. New Olef's Italian-English meetings became fully searchable in both languages.
For manufacturing companies evaluating AI meeting tools, the question is not whether the technology works. It is whether the tool can handle the specific vocabulary, cross-departmental complexity, and compliance requirements that define industrial operations. New Olef's experience suggests that when those boxes are checked, the efficiency gains are substantial and measurable.