See what your data is doing.
Sixty-four seconds inside Marver Workbench: schema authoring, self-refreshing workspaces, and identity resolution, captured live.
The platform that should have existed.
Live data collection, user-defined schemas, cross-channel identity resolution, and a desktop analyst workbench, running on infrastructure your team controls.
A real desktop workbench
A desktop app for the analysts who lived in Data Workbench. Explore paths, funnels and correlations interactively against your own server, not in a browser tab waiting on a cloud query queue.
Model your business
Define schemas that match the way your business actually works. Add fields and dimensions without a migration or a consultant; only breaking changes need a rebuild.
Unify the same customer Analytics + CDP
Resolve identity across web, server-side and batch-file sources with deterministic matching, a full history of how each profile formed, and un-merge when a link is wrong.
Fresh data, not yesterday's batch
New events are queryable within about a minute at default settings, and open workspaces refresh themselves while you watch.
Blend every channel into a single timeline for each customer.
Data Workbench was built to analyze customer interactions online and offline in one place. Marver keeps that model: you route live digital events and imported files into the datasets you define, and rows that share an email, customer ID or user ID resolve to the same customer before they get there.
Live digital channels
Web and Next.js SDKs, events forwarded from Google Tag Manager or Tealium, and any system that can POST JSON over HTTP.
Offline and batch sources
CSV, JSON and NDJSON files from CRM, point-of-sale or call-center systems. Each is a registered source with its own join key and consent purpose, and every import returns a receipt of rejected and consent-suppressed rows.
Reference data
Lookup dimensions that turn raw codes into business names through a direct mapping, a pattern, or an uploaded CSV: SKU to product family, store ID to region.
All your channels, one profile per customer Analytics + CDP
Route your channels into a shared dataset and Marver links each customer's web visits, store receipts and support calls to a single profile, while every row keeps the channel it came from. Compare channels side by side, or follow one customer across all of them in Path Browser, Funnels and attribution.
Evaluate it on your hardware, on your schedule.
Request a license, start the server with Docker Compose, and point real traffic at it. Your customer data never touches our systems, and every question gets a written answer from the engineer who built it.
Inside your network
Your channels
Web and tag-manager events, server-side calls, and CRM, POS or call-center files
Marver server
Collection, processing and identity, on your hardware or cloud account
Marver Workbench
Desktop app on your analysts' machines: macOS, Windows and Linux.
Leaves your network
A license check, once a day
A license ID and an install fingerprint. Never customer or event data. It can go through your HTTP proxy. If it can't get through for 30 days, Marver goes read-only until it can. Air-gapped Enterprise licenses skip it entirely.
Request a license
Tell me your company and where the server will run. I reply with a signed 45-day license file and the Docker Compose bundle.
Run it on your infrastructure
Bring the server up on a Linux host (x86-64 or ARM64), or under Docker Desktop on a Mac or Windows machine. Analysts install the desktop Workbench and connect to it.
Keep it, or let it lapse
To continue, reply for an order form or invoice and drop in the paid license, with no reinstall. If you don't, Marver stops taking new events and you keep view, query and export access to everything it already holds. Raw events sit on your disks as plain NDJSON text, readable without Marver.
Best for teams with a real data ownership problem.
Marver is a strong fit when customer data control, identity quality, and analyst workflow matter more than buying another generic dashboard.
Best fit for
Probably not for
Marver for your team: Data Workbench teams · Healthcare · Financial services · Public sector and EU · Mid-size B2B
Pricing that scales with your deployment.
Not with your headcount.
No per-seat fees, no MAU pricing, and no event metering on paid tiers. Start with a full 45-day trial on your own hardware, then pick the tier that fits. Every tier includes unlimited users.
Try everything free for 45 days: the full platform, including CDP identity resolution, on your hardware. No card required, and no customer data leaves your network.
Analytics
$2,500
per month · flat
Unmetered events · No identity layer
The analyst workbench with no event caps: schema authoring, Path Browser, Funnels, Correlation and cluster analysis, with role-based access for every user.
Analytics + CDP
$5,000
per month · flat
Persistent customer identity
For teams that need trusted customer identity, not just event counts. Everything in Analytics, plus unified profiles, attribution, an identity formation audit with un-merge, and direct founder support.
Enterprise
Custom
based on scale
Air-gapped · Custom terms
Everything in Analytics + CDP, plus offline licenses for air-gapped networks, a TLS-on deployment profile, custom license terms, and help with your security review. SSO and high availability are on the roadmap.
Add the intern. Add the VP. Add the auditor. Add the data science team that rotates every six months. The price doesn't move.
Security questionnaires, order forms and purchase orders are handled by email.
Download Marver 1.2.0
The Workbench installs on your own machine in a minute. First launch opens the live public demo on its own, so you can click through real data before asking for a license.
Released November 2, 2026
macOS
Not sure which Mac you have? Open the Apple menu and choose About This Mac: a Chip line means Apple Silicon, a Processor line means Intel.
Linux
Use the .deb on Debian and Ubuntu. The AppImage runs on most other distributions.
SHA-256 checksums
- Marver-1.2.0-arm64.dmg
- 58222e567a75b4fed8f467a4d94151c08f654c6961bc04a4ddbf3791051714fa
- Marver-1.2.0.dmg
- 456d5b05e12fd6f11d715bb798e35577c546f66e3b918a91a8405ce167110211
- Marver.Setup.1.2.0.exe
- c91895e098a1beaa51c2593a3b7b8d2b9cf35763b0c805e8c505e7af57028af3
- Marver-1.2.0.AppImage
- a770d95e0ac15b6f71c79e3b98d701d2c7002c132005f1c1151b557cd0b9183e
- marver-workbench_1.2.0_amd64.deb
- c69e24534b08794ca8322954e0bd7759c8aee8a90aed9cc2821221ddb8767b3f
- Marver-1.2.0-arm64.AppImage
- 8f592b7f0a8cfa378ec7a4519f997eb9760d4f2216ca3b7952cf219e43cd4110
- marver-workbench_1.2.0_arm64.deb
- ba4f323c978c8797c6b9d46d043e8dc9e9a4e0f176de017c8ca5dd8ef0507dce
1. Install
Run the installer and launch Marver.
2. Explore the live demo
First launch opens the live public demo and walks through it in four steps. No license, no email.
3. Run your own data
Request the 45-day trial below to get a signed license.dwb and the server bundle.
The questions your CISO will ask, answered by architecture.
Self-hosted architecture does not replace legal review, but it does give security, privacy, and procurement teams cleaner answers about where data lives, who controls it, and which vendors sit in the path.
Data residency
Keep customer data inside infrastructure and jurisdictions your team controls.
Vendor exposure
Reduce dependence on third-party analytics vendors, subprocessors, and external data paths.
Sensitive workloads
Run regulated workloads with TLS and mutual TLS between components, three-role access control, and an audit trail of configuration and identity changes.
Air-gap capable
Enterprise licenses can run fully offline, with no license server and no outbound calls.
Deletion and retention
Retention and deletion happen on storage you own. No vendor can read, copy, or delete your customer data.
Audit posture
Make it easier to explain where customer data lives, who can access it, and how it is governed.
Solved by architecture · Not by contract
The only person building this is the same person who had to buy it.
Andrew Chepurny
Founder & Principal Architect, Chip Hosting Solutions
I built and operated Adobe Data Workbench at enterprise scale for Fortune 500 organizations. When Adobe announced end-of-life, three clients called me in 48 hours. Telecom. Insurance. Financial services. All asking the same question:
"Where do we go now?"
Everything was cloud-only. The legacy thick-client power-analyst workflow (the one data scientists actually used to run path analysis, attribution, and statistical exploration) just disappeared. I didn't have an answer for them then. So I spent the last 5 years engineering one.
Analytics architecture delivered for
Consulting and in-house engagements, not Marver deployments.
Built and operated Adobe Data Workbench at enterprise scale. Clients include Dollar General, Amtrak, Western Union, ForeFlight, Pfizer, City National Bank, and Adobe Systems, plus many more.
You are not buying a black box. You are buying the platform and written answers from the person who built it.
Adobe where it fits. Marver where the data can't leave your network.
Adobe consulting from Chip Hosting SolutionsDeploy on bare metal or your own private cloud.
Some teams have the regulatory headroom to deploy on cloud infrastructure they fully control. If that's you, nothing stops you: your AWS account, your Azure tenant, your GCP project, or a local Linux server. The architecture doesn't care where it lives.
What matters is that you own the keys, you own the data, and you own the machine.
The difference isn't cloud versus on-prem. It's renting versus owning.
Request your 45-day trial.
Tell me who you are and where the server will run. I review each request myself and reply with a signed license.dwb file and the matching install bundle, usually within one business day.
Prefer email? Write to andrew@chip-hosting.com
Licensing question? Use the contact form · Implementation partner? Talk partnerships
License commitment
Marver has been commercially licensed since day one. No BSL relicensing. No OSS-to-proprietary rug pulls. No fork risk. Ever.
© 2026 Chip Hosting Solutions · The product name is Marver.