Logistics — AI-driven adaptive learning & ERP training
A proprietary ERP patched together over years — a powerful system and a terrible teacher — with a workforce of high-turnover drivers, rotating dispatch shifts, and quarterly sales onboarding all using it differently.
The workforce was all over the map: drivers with high turnover, dispatch coordinators on rotating shifts, sales reps onboarding every quarter, and a small technical team keeping the platform alive. Everyone touched the ERP differently, and almost nobody had a clean path to learn it. Three problems kept showing up.
ERP complexity at the field level. Drivers and floor staff struggled with fueling and fleet modules — wrong entries, missed steps, duplicate logs. A bad fuel-stop record didn't stay a training issue; it showed up as dispatch delays, reconciliation headaches, and compliance flags downstream.
Training ate senior time. Experienced sales managers and ops leads hand-held new hires through playbooks and walkthroughs for hours every week — time taken directly out of revenue work and route planning.
One manual for everyone. Static PDFs and slide decks treated every role and pace the same. Sales reps skimmed, drivers never opened them, technical staff needed depth the manuals didn't have. Engagement was low and performance inconsistent across depots.
We built a custom AI-augmented learning platform that worked as a 24/7 tutor — role-aware, source-backed, and tied to how the company actually operated.
We indexed the full internal library — ERP manuals, module guides, sales playbooks, safety protocols, dispatch SOPs — into a retrieval-backed Q&A layer. Employees could ask plain questions like "How do I log a fuel stop?" and get step-by-step answers with source references. If the answer wasn't in the knowledge base, the system said so — which mattered on compliance-sensitive workflows.
Different curriculum tracks by function: sales got outreach simulations and CRM/playbook drills tied to their real motion; drivers got interactive ERP navigation and fueling-compliance walkthroughs on mobile-friendly modules; technical staff got deeper maintenance, troubleshooting, and escalation paths. Same platform — different entry point, depth, and certification bar.
The system tracked where each user struggled in real time — failed items, repeated help requests, slow steps — and adjusted difficulty, surfacing extra practice when someone stalled on fuel logging or dispatch handoffs. Fast movers weren't forced through remedial slides; people who needed reps got them without waiting for a manager to notice.
Tests were generated from the same source material, scored instantly with qualitative feedback on gaps rather than a bare pass/fail. Leadership got a readiness dashboard — completion by role, module-level weak spots, certification progress, and heatmaps of where live-ERP errors still clustered after training. The loop closed between what we taught and what still broke in production.
| Metric | Before | After |
|---|---|---|
| Median time to operational proficiency (new hires) | ~4.5 wks | ~1.3 wks |
| ERP user-error reports (fleet & fueling) | baseline | −90% |
| Internal ERP how-to support tickets | 140 / mo | 28 / mo |
| Senior-manager hours on manual onboarding | ~18 hrs/wk | ~3 hrs/wk |
| Fuel-stop logging accuracy (audit sample) | 81% | 96% |
| Module completion rate (role-assigned paths) | 52% | 89% |
| Employee confidence in role tooling (survey) | — | 85% higher |
Onboarding shifted from a manager-led bottleneck to a mostly self-serve path with checkpoints. Drivers cleared ERP certification before their first solo route instead of learning through mistakes, and training became something people actually used mid-shift when they hit an edge case — not a one-time orientation event.
Early answers pulled from outdated manuals. The knowledge base launched before every document was reconciled to the current ERP version. For about ten days the tutor confidently explained a fuel-logging flow replaced two releases ago, and driver trust dropped hard until we added version tags and a "last verified" stamp on every answer.
Drivers needed it on the road, not at a desk. The first build assumed stable connectivity and a full browser session. Load times at fuel stops and patchy signal made it frustrating, so we shipped lighter mobile modules and cached the highest-frequency guides offline. Driver usage moved only after that.
Adaptive pacing was too aggressive at launch. The algorithm advanced people quickly to keep engagement up, and assessment failures spiked on dispatch and compliance modules — people passed screens without retaining steps. We slowed progression and added mandatory practical checkpoints.
Sales simulations felt fake until we rewrote them. Generic role-play got ignored as "training homework." Once scenarios mirrored real objection patterns and account types from their playbooks, completion and scores improved together — the content layer mattered as much as the AI layer.
Readiness heatmaps landed wrong with middle management. Ops leads read team weak spots as surveillance, and two depot managers stopped assigning modules. Reframing the dashboard around "where to coach" — and giving leads control of remediation — got adoption back.
Tribal knowledge wasn't in any manual. Senior dispatchers knew workarounds the docs never captured, and RAG can't retrieve what was never written down. We ran structured capture sessions with tenured staff and fed those into the knowledge base.
- Training for ERP adoption has to live in the language people use. Drivers think in fuel stops, load checks, and handoffs — not "modules." Tutorials framed in their terms got used; manual language didn't.
- Source-verified answers are non-negotiable on operational systems. A wrong training answer in logistics isn't embarrassing — it's expensive. Citations, version control, and explicit "I don't know" built more trust than smoother prose.
- Adaptive learning needs guardrails, not just speed. Personalization works when it catches struggle early; it fails when it optimizes for completion over retained competence.
- Role-specific paths aren't cosmetic. Sales, drivers, and technical staff needed different outcomes, certification criteria, and practice environments — not different skins on one course.
- Automating training doesn't remove managers — it changes their job. The 15+ hours reclaimed weekly shifted from repeating basics to coaching on exceptions and account strategy.
- Capture tribal knowledge before you scale the engine. RAG is only as good as what's indexed, and the last mile of training content often lives in people's heads until you go get it.