Insights · Planned

Device issue reports and AI rollout reports

Tracebacks from your devices, grouped into issues, and rollouts that report back in plain English. Opt-in, run on our servers, at no cost to the device's battery.
Device issue reporting

Crash reporting for MicroPython devices.

Like Sentry, for microcontrollers: unhandled exceptions and the ones your code reports become tracebacks, grouped into issues across the fleet, with the device diagnostics that explain them.
Planned
Grouped by type and message, on the device first, then across the fleet
With firmware and app version, battery, signal, uptime and chip
New and regressed issues alert in the dashboard, Discord and webhooks
Rate-limited on the device, so a crash loop can't flood a check-in
0
extra wake-ups or radio time: reports ride along with the next check-in
Like Sentry, for microcontrollers: tracebacks grouped into issues, sent with the next check-in at no extra radio time.
Regression detection and rollout reports

Rollouts that report back.

Staged rollouts with AI regression detection: updated devices are compared with those still on the old release, and a wave that looks worse pauses, with a plain-English explanation of why.
Planned

What it compares

Compares updated devices with those still on the old release (new issues, reboot loops, failed check-ins, battery drain, signal) and pauses a wave that looks worse, with an explanation of why.

What it reports

After each wave and at the end: what changed, what updated, rolled back or failed, which issues came or went, and the battery and check-in impact, in plain English to the dashboard, Discord or email.
A statistical test decides when a wave pauses, and people approve releases. The model only explains what the numbers show.
Anomaly reports

Device health monitoring, between rollouts too.

Watches the fleet all the time, not only during rollouts, and flags faster battery drain, check-ins stopping at a site, reboot spikes, signal dropping on one network, or a device checking in more often than scheduled.
Planned
Release impact reports, on the battery page ›
Package change reports (Planned) show which packages each release changes and what each does to battery life.
Package change reports ›
Device battery and power status

Every battery in the fleet, at a glance.

Every device's battery level and power status (on battery, on fixed power, or charging), a battery-level histogram for the fleet, an estimate of when each battery runs out where the data supports one, and low-battery alerts. Devices with a backup battery report power lost and restored; when several without one go silent together, otalite flags a possible outage, installs nothing during it, and doesn't count those devices as rollout failures.
Planned
A run-out estimate appears only where a device's own history supports one.
Triage and release notes

Triage, release notes and windows, drafted.

Suggestions you check and accept, never changes made for you.

Issue triage

Groups similar tracebacks, suggests a likely cause, and points to the probable commit.
Example
Likely cause: the retry added in a41c9e2 reconnects Wi-Fi on every failed upload.
Planned

Release notes from commits

A draft written from the commits between two releases, for you to edit.
Example draft
thermo 1.4.2 • Uploads retry after a dropped connection • Humidity read every 10 minutes
Planned

Smarter update windows

Suggests an update window from your fleet's real check-in and battery patterns.
Example
Suggested: Sunday 03:00–05:00, when 92% of Berlin devices are awake and above 50% battery.
Planned
Ask your fleet

Questions in plain language.

In the dashboard, or with /otalite ask in Discord. Still being researched: whether and how it can answer reliably.
Exploring
Trust

Opt-in. On our servers. Not for training.

The analysis runs on our servers, from data devices already send with their check-ins, so it costs the device's battery nothing.
Recommended
AI features are opt-in per organization and use only rollout metrics, issue reports and tracebacks. Customer data is not used to train models.
Lighter option
Let it read source diffs from your connected repository for sharper root-cause suggestions. That code is then processed by a third-party AI model provider, and you accept that risk.
Status

Where each piece stands.

Nothing here is available yet. In development is being built for the first release; Planned is on the plan, not started; Exploring is still being researched.
Observability

Device issue reporting

Like Sentry, for microcontrollers: tracebacks grouped into issues, sent with the next check-in.
Planned
Energy

Device battery and power status

Each device's battery level and power status, a fleet histogram, run-out estimates, and low-battery alerts.
Planned
AI-assisted rollouts

Regression detection during rollouts

Updated devices compared with the old release; a wave that looks worse pauses, with the reason.
Planned
AI-assisted rollouts

Rollout reports

After each wave: what updated, rolled back or failed, and the battery impact, in plain English.
Planned
AI-assisted rollouts

Anomaly reports

Watches the fleet between rollouts too: faster drain, silent sites, reboot spikes, weak signal.
Planned
AI-assisted rollouts

Issue triage

Groups similar tracebacks, suggests a likely cause, and points to the probable commit.
Planned
AI-assisted rollouts

Release notes from commits

A draft written from the commits between two releases, for you to edit.
Planned
AI-assisted rollouts

Smarter update windows

Suggests an update window from your fleet's real check-in and battery patterns.
Planned
AI-assisted rollouts

Ask your fleet

Ask in plain language, in the dashboard or with /otalite ask in Discord.
Exploring
AI-assisted rollouts

Release impact reports

Each release compared with the one before, for battery, CPU and network use.
Planned
See the whole roadmap ›
Related

Keep exploring.

Rollouts

Rollouts that stop themselves.

Waves that widen only when it's safe, and pause on failures.
Learn more ›
Battery

Built for devices that sleep.

Schedules, battery targets, and months of battery simulated in minutes.
Learn more ›
Workflow

Ship from where you already work.

Git, CI, the CLI, your phone, Discord and your coding agent.
Learn more ›
Community

Follow the build.

otalite isn't available yet. Register your interest to hear first, or join the Discord to see what we're building and tell us what your fleet needs.
Register interest
Join the Discord
See the roadmap
© 2026 otalite · Coming soon
FeaturesBatterySecurityRolloutsInsightsWorkflowRoadmapMicroPython OTA optionsBattery life calculatorCyber Resilience ActAboutRegister interestDiscordGitHub