secure by default
Analyze logs locally —
offline, private, nothing uploaded
Your logs stay yours.
Logs often contain the exact things you don't want floating around the internet — user data, tokens, internal hostnames. Logier is built so that reading them doesn't require handing them to anyone.
a different starting point
Logier vs. a typical cloud
log platform
| Logier | Typical cloud platform | |
|---|---|---|
| Where logs are processed | On your machine, in memory | Uploaded to a vendor's servers |
| Setup required | Open a file or paste text | Agents, ingestion pipelines, schemas |
| Data retention | Only for the session you're in | Indexed and stored, often for weeks |
| Account needed to analyze a file | No | Usually yes |
| Internet connection required | No, for parsing and analysis | Yes |
data-handling principles
How Logier handles
your log data
In-memory by default
Parsing, clustering, timeline building, and anomaly detection all run locally against the file or pasted text you gave Logier, and that data lives in memory for your session. Nothing is written to a remote server as a side effect of simply opening a log.
Nothing leaves without an action
The features that talk to an external service — AI root cause analysis and Live Mode — are both opt-in: Copilot sends only the excerpt you're asking about to the provider you configured, and Live Mode only connects to the URL you give it. Every other feature works entirely offline.
Sessions, not archives
Logier is built for the debugging session in front of you, not as a system of record. When you close a file, Logier isn't holding onto a permanent copy of it somewhere you'd have to remember to go delete.
Local files stay local files
Dropping in a file from disk doesn't mean uploading it anywhere first. Logier reads it directly from your filesystem, the same way any other desktop application would.
what this means in practice
Analyzing logs that contain
PII, credentials, or customer data
You can open a log containing customer data, internal URLs, or partial credentials without a second thought about where that content ends up. You can debug on a plane, on a network you don't trust, or on a file you're not allowed to email to anyone outside your team — because for Logier's core features, there's nowhere for that file to go except the memory of the machine it's already sitting on.
offline & air-gapped
Works fully offline and
on air-gapped machines
Logier is a desktop application, not a client for a hosted service. Pull the network cable and everything below still works, because none of it was ever going over the network in the first place.
Opening and parsing files
Format detection, parsing, and the fallback parser are all local pattern matching.
Search, filters, and clustering
Everything that narrows a file down runs in memory on your machine.
Timeline and anomaly detection
Both are computed from the file you opened — no baseline is fetched from anywhere.
No account, no licence phone-home
You don't sign in to read a log, and the app doesn't need to reach a server to let you work.
The one feature that needs a network is the optional AI analysis — and even that can be pointed at a local model running on the same machine, so nothing crosses the boundary. See exactly what it sends.
where this lives
Everything here runs on your machine
There is no ingestion step between opening a file and reading it. What you see below is a local application reading local files — no account, no upload, no indexing service in between.
Explorer
Cloud
- 1
Files are read directly from your filesystem, the same way any desktop application would.
- 2
Parsing and clustering happen in memory, for this session only. Nothing is written to a remote server.
- 3
The chart and the anomaly baseline are computed locally from the file itself — no model, no vendor.
- 4
The only feature that uses the network, and only with a provider you configured — or a local model.
why logs specifically
Logs are the least sanitized data you own.
Production logs routinely contain session tokens, email addresses, internal hostnames, full request bodies, and stack traces naming internal paths — captured incidentally, by code written under deadline, and never reviewed for what it emits. Uploading that to a third-party ingestion pipeline to debug a single crash is a disproportionate transfer, and in a regulated environment it may not be available to you at all. Logier processes the file in memory on your machine: no account, no ingestion endpoint, no background upload.
faq
Common questions
keep reading
What runs locally
AI root cause analysis
The full data boundary: what the excerpt contains, and how to run it against a local model instead.
Automatic format detection
Format detection is local pattern matching — your file is never uploaded to identify it.
Anomaly detection
Statistics computed on your machine from your own file — no training data, no model, no vendor.
Unknown log format analyzer
Proprietary software tends to mean sensitive logs. Here's how Logier reads formats nobody has documented.
Offline log viewer
Air-gapped machines, customer log bundles, and working with no connection at all.
Logier vs. cloud log platforms
When ingestion, alerting and retention are genuinely worth it — and when they aren't.
Debug without giving anything up.
Local by default, with exactly two clearly-marked, opt-in exceptions.