UA/EN
Safety + AI2026

Aeris

Making the invisible airspace visible.

A civilian early-warning system. It fuses official alerts and open sources into zone-level risk — with probability, confidence and an explanation.

Round
€400k · 18 mo
Stage
v1.0 in production
Platforms
Web, API; iOS and macOS not yet released
For
Public, organisations
Coverage
All oblasts of Ukraine
Aeris

(01) Problem

Threat information is scattered across official alerts, Telegram channels and monitoring communities. Reports duplicate each other and nobody rates how reliable they are. People and businesses struggle to see what is happening now and what comes next.

(02) Solution

Aeris gathers signals from many sources, removes duplicates and shows risk for every zone at 5, 15 and 30 minutes. Every assessment comes with a plain-language explanation: why the system thinks so and how confident it is.

(03) How it works

  1. 01

    Collect data: official alerts, open sources, weather and, later, partner sensors.

  2. 02

    Normalise everything into one format and map it to zones, never exact coordinates.

  3. 03

    Reports from one origin count once: 100 reposts don't raise confidence.

  4. 04

    ML models score zone risk overall and for drones, missiles and glide bombs.

  5. 05

    People see the result on the lock screen, in the Dynamic Island, in the Mac notch or in the Control Center.

(04) Features

iOS app

Map and list of oblasts, alert start/end notifications, a timer on the lock screen and in the Dynamic Island. No sign-up.

macOS notch app

Alert status next to the MacBook notch, with a sound when an alert starts.

Control Center

An operator console: risk map, incidents, forecasts, anomalies, infrastructure, sources.

Public alerts board

A region map at aeris-app.com/alerts.

API

REST and WebSocket for integrations, a public alerts endpoint.

Explanations

Every assessment has probability, confidence, data quality and reasons.

(05) Business model

The public app is free — it builds reach and trust. The paid side is the platform for organisations: Control Center, API and sensor integrations. Pricing for organisations is not set yet.

Public appFreeiOS, macOS, alerts board
OrganisationsOn requestControl Center, roles, multiple teams, forecasts, infrastructure
IntegrationsOn requestAPI, partner cameras and acoustic sensors

(06) Status

Version 1.0 has been in production since 1 October 2026: public site, alerts board, API and Control Center. Official alert data has been flowing since 12 October. The iOS app is not yet submitted to the App Store; it will show the official alert status by oblast, while risk forecasts are a feature of the platform for organisations.

17/17platform phases done
3 553events in 30 min, zero errors
6 msdata to publish (p95)
650+automated tests
2 641Kyiv alerts in history since 2022

(07) Market

Ukraine has the world's most battle-tested public-warning audience: one alert app alone has 46M downloads and 6.7M daily users. Mass notification is forecast at ~$47B by 2030, and Everbridge's $1.8B buyout shows what a critical-event-management platform can be worth.

46M“Air Alert” app downloads
6.7M+daily users of that app
14 870air-raid alerts in Ukraine in 2022 alone
~$47Bmass notification market, 2030 forecast

Competitors

“Air Alert” (Ajax), alerts.in.ua, Kyiv Digital — show the official alert status. Everbridge, Genasys — enterprise and government notification. Aeris differs with multi-source fusion, 5–30 minute zone risk forecasts with explanations and a platform for organisations, not just a siren on the phone.

Sources

  1. Ajax Systems, «Повітряна тривога» — 4 роки, Mar 2026 ↗
  2. Ukrinform, 14,870 air-raid sirens in 2022 ↗
  3. Grand View Research via Security World Market, 2025 ↗
  4. Everbridge 8-K, Thoma Bravo acquisition, 2024 ↗
  5. MarketsandMarkets, Critical Infrastructure Protection ↗

(08) Technology

  • Source independence: related reports merge into one, confidence grows only from independent confirmations.
  • ML: clustering, anomaly detection, zone forecasts, calibrated probabilities, model drift monitoring.
  • New data sources plug in via configuration, no code; tested with 120 sources.
  • Raw data is archived unchanged and can be reprocessed with newer algorithms.
  • Java 21 / Spring Boot, PostgreSQL + PostGIS, Python ML service, Next.js, SwiftUI.
  • Ethical limits in code: no targeting, no exact object coordinates, trajectories or tracking of people.

(09) Roadmap

Oct 2026Platform v1.0 in production, official alerts, Control Center■ Done
NextiOS app in the App Store□
NextPublic macOS release and lock-screen widget□
NextPartner data source integrations□
NextCitizen sensor mode on the phone — opt-in, processed on device□

(10) Screens

(11) Round

€400kround size
18 morunway

Raising €400k for 18 months — investment, grants or a mix. The platform already runs in production; the money goes into releasing the apps, new data sources and the first paid pilots with organisations.

Pre-seed, SAFE · can be combined with grants

Use of funds

EngineeringiOS and macOS, widgets, risk ML models40%
Data & sensorsPartner sources, acoustic sensors, citizen sensor20%
Pilots & salesControl Center for communities and infrastructure operators15%
Security & auditExternal audit, attack resilience10%
Contingency15%

What it delivers (targets)

300kpublic app installs
3paid pilots with organisations
€150kARR from organisations
2apps: iOS and macOS

What's next

Next, a seed round to support contracts with government and infrastructure organisations in Ukraine and EU countries facing similar threats.

Want to know more about this product? We'll send metrics, the financial model and a demo.

Next product → Secret Places