Case Studies

Real Projects. Measured Outcomes.

Six detailed engagements — challenge, solution, result — plus what the client said afterward. No vanity metrics, no stock photography.

Manufacturing · 2024

Northwind: 42% Less Unplanned Downtime Across 4 Plants

Challenge: 340 CNC machines across 4 plants, each with a different PLC vintage. Maintenance was reactive; a single bearing failure could cost $180K in lost production.

Solution: 6-week pilot on 2 asset classes → vibration + thermal anomaly model trained on 18 months of historian data → CMMS integration for auto-generated work orders → scaled to all 340 assets by week 16.

Result: 42% reduction in unplanned downtime, $1.7M avoided cost in year one, and a maintenance team that now spends 60% of its time on planned work.

-42%
Unplanned downtime
$1.7M
Saved / yr
16 wk
Full rollout
Daniel Reyes★★★★★

"AutomataBridge rebuilt our OT network in six weeks and cut a ransomware blast radius from 14 plants to 1. Their engineers speak both PLC and Kubernetes — rare."

VP Operations · Northwind Manufacturing
Energy · 2023

Helios: Zero-Trust OT Across 14 Generation Sites

Challenge: 14 generation facilities with flat OT networks, shared credentials, and an IEC 62443 audit scheduled in 90 days. Previous MSP had flagged zero findings — the board was skeptical.

Solution: IEC 62443 gap assessment → Purdue-model segmentation → microsegmentation with OT-aware firewalls → 24/7 MDR onboarding → tabletop incident-response exercise.

Result: Blast radius reduced from 14 plants to 1, MTTR from 3 hours to 11 minutes, and IEC 62443 audit passed on the first attempt.

-93%
Blast radius
11 min
MTTR
100%
Audit pass
Amelia Kowalski★★★★★

"We went from 3-hour incident response to under 11 minutes. The SOC is genuinely 24/7 — not a voicemail box at 3 a.m."

CISO · Helios Energy
Healthcare · 2024

MedAxis: Clinical AI Copilot With 92% Adoption

Challenge: Clinicians spent 40% of their day on documentation. Previous AI vendor pilot was abandoned after 3 weeks — hallucinations and no audit trail.

Solution: HIPAA-safe RAG pipeline on 2.1M clinical documents, private LLM instance inside MedAxis's tenancy, eval harness with 1,200 clinician-reviewed test cases, human-in-the-loop review on every output.

Result: 58% reduction in documentation time, 92% clinician adoption at 90 days, and a full audit trail that satisfied the compliance team on first review.

-58%
Doc time
92%
Adoption
1,200
Eval cases
Dr. Jamal Price★★★★★

"They embedded an AI copilot into our clinical documentation in 90 days. Clinicians actually use it — that's the bar, and they cleared it."

CMIO · MedAxis Health
Financial Services · 2023

Cobalt: Azure Migration, Zero Downtime, 31% Cloud Savings

Challenge: 140 on-prem VMs, an expiring data-center lease, a 9-month deadline, and a FinOps bill nobody understood.

Solution: Landing zone in 6 weeks → 8-wave migration with zero downtime windows → Kubernetes re-platform of 22 services → FinOps dashboard with tagging enforcement and rightsizing.

Result: 100% of workloads migrated before the lease expired, zero customer-facing downtime, and a 31% reduction in monthly cloud spend by month 4.

-31%
Cloud spend
0 min
Downtime
140
VMs migrated
Sofia Lindqvist★★★★★

"Migration to Azure was the least painful project of my career. Zero downtime, honest budgets, and their FinOps review saved 31% on our monthly cloud bill."

CTO · Cobalt Financial
Logistics · 2022

Vertex: Fleet Telemetry Consolidated Across 6 Vendors

Challenge: 1,800 vehicles, six different telematics vendors, and a dispatch team manually copy-pasting between portals. Nothing reconciled.

Solution: Unified telemetry pipeline (MQTT + Kafka) → normalized vehicle events → dispatch automation rules → real-time ops dashboard on edge nodes at 4 hubs.

Result: Dispatch time cut by 37%, $1.7M in avoided bearing failures via predictive maintenance, and a single source of truth for fleet operations.

-37%
Dispatch time
$1.7M
Failures avoided
1,800
Vehicles
Marcus Tan★★★★★

"Their predictive-maintenance model saved us $1.7M in avoided bearing failures in the first year. ROI was obvious by month four."

Director of Reliability · Vertex Logistics
Telecom · 2024

Orbital: Edge Compute Deployment Across 220 Sites

Challenge: A 5G rollout required edge compute at 220 tower sites within 9 months — with no central management plane and no field team on standby.

Solution: Standardized edge node image → remote OTA fleet management → zero-touch provisioning → 24/7 NOC integration with Orbital's existing ticketing system.

Result: All 220 sites deployed on schedule, 99.7% node uptime in the first quarter, and no on-site visits required after initial installation.

220
Sites
99.7%
Node uptime
0
Post-install visits
Robert Bell★★★★★

"Their CEO personally jumped on a P1 call on a Sunday. That told me everything about the company culture."

VP IT · Orbital Telecom
Engineering Newsletter

Monthly Technical Deep-Dives

Subscribe for monthly technical deep-dives on predictive maintenance, edge computing, and OT security. Written by our engineering team.

No spam. Unsubscribe anytime. We never share your address.