Blue Mesh
Talk to us
Customers

Case studies

Blue Mesh in production, told with the figures each customer reports in its own case study, and the organizations that run it and agreed to be named.

Talk to us →See the platform
Trusted in production

The customers we name publicly

Each of these organizations runs Blue Mesh and agreed in writing to be named. Netstratum and Voifinity are Netstratum-family brands: real deployments, not evidence of adoption outside the family.

  • 01Vaspian
  • 02Lonesome No More
  • 03Voifinity
  • 04Netstratum

The problem

Every customer conversation carries valuable intelligence and sensitive personal data. Vaspian saw the opportunity early: thousands of daily calls held rich insight on customers, sentiment and service quality that its traditional stack could not capture. Unlocking it meant AI that could operate on regulated communications data without compromising customer privacy, at the volume of a working telecom network rather than a pilot.

What we did

Vaspian chose the Blue Mesh Intelligence Layer, a speech-analytics and conversation-intelligence platform over its call recordings: AI transcription enriched with entity detection, speaker diarization, sentiment and intent, with personal data redacted automatically.

Results

  • 01Around 950,000 AI transcript requests processed every month
  • 02Around 38,000 hours of audio transcribed and analysed
  • 03Customer PII protected through automated redaction
  • 04Consistently high availability under sustained enterprise load

The problem

Lonesome No More set out to build an AI voice companion for seniors, a service whose conversations touch health information and deeply personal detail. The team evaluated the leading AI voice agents and, by its own account, found none that combined natural conversation with the privacy such data demands, so public, shared AI platforms were not an option. They needed a voice agent that feels human, on infrastructure private enough to be trusted, and one that could hold a conversation with an older person on an ordinary phone line: hearing difficulty, slower speech, long pauses that are not the end of a turn, and no screen to fall back on when the model gets it wrong. The case study sizes the market at 24 million seniors.

What we did

The answer was a custom-trained voice model in a dedicated private Blue Mesh instance, governed by safety guardrails, with conversations held inside it. Connected directly to the telephone network, the agents hold natural daily check-ins, remember prior conversations for continuity, and monitor wellbeing, escalating concerns to carers with a person deciding what happens next. Software only, no hardware in the home.

Results

  • 01Reliable daily automated engagement
  • 02Up to 60% cheaper than hardware-based alternatives
  • 03After the initial capital investment, ongoing running cost stays low
  • 04Health data and personal details are held inside the private instance
  • 05Conversation history retained for continuity, with concerns escalated to a person who decides what happens next
Before you decide

What to check before a deployment

How an evaluation works

One working session on a workflow you already run, on a sample of your material, with your team holding the gate. What we need from you, and how pricing is set afterwards.

Ask for an evaluation →

Where it runs and who is in control

Private cloud, on-premises or fully air-gapped. Role-based access, escalation to a person on the decisions that matter, and an audit trail your reviewer can read.

Trust and security →

How Blue Mesh compares

Every platform on your shortlist held beside ours, cell for cell, from what each vendor publishes, with the questions worth putting to all of them, including us.

How we compare →
The landscape

Compare us with any platform

An enterprise AI platform is a multi-year commitment. It will sit inside your environment, act in your systems, and answer to your auditors, so it should be chosen on the criteria that decide those things, not on the demo. Our own published answers come first; any platform's answers can be held beside them, cell for cell, from each vendor's own pages.

PlatformDeploymentMulti-modelVoiceHuman approval before an actionPublished compliance
Blue Meshthis is usPrivate cloud, on-premises, or fully air-gapped. BM in a Box is an air-gapped appliance on NVIDIA DGX. sourceYes sourceYes, over SIP sourceYes, every level sourceCertifications your build can carry: SOC 2, ISO 27001, HIPAA. Every decision logged, role-based access, encryption. source

1. Built from what each vendor publishes on its own website, with the source linked in every cell. Last updated 18 August 2026.

2. All product names and trademarks are the property of their respective owners; no affiliation or endorsement is implied.

Where to go next
  • Voice agents Behind the Lonesome No More companion: low-latency voice over SIP, speech in and out, running on models you host.
  • Telecom Vaspian's sector: call recordings transcribed and read for sentiment, quality and compliance risk.
  • Platform overview The modules, capabilities and deployment modes behind both deployments, on one page.
// customers

Bring a workflow you already run

In one session we will build an agent for it, live, with your team holding the gate, so you judge the work rather than the references.

Talk to us →