Finland's security situation is written about every day in official bulletins, the news, research institute reports and social media. Putting those together into one picture is hard to do alongside other work, and some of the information arrives already coloured by somebody's narrative.
In a company the same collecting is often done several times over: for a client, for an internal briefing and for the management's question about whether anything has changed. HÄIVE collects, classifies and checks the material once and keeps it up to date. Your own time is left for interpretation and client work.
The service runs in a browser on a phone or a computer, in Finnish and in English. There is nothing to install, and an email address is enough to sign in. The front page shows in a couple of minutes whether anything has changed, and every topic has a page of its own for those who want the reasoning.
A news story tells you what somebody has already noticed. Every evening HÄIVE also reads measurement data that newsrooms rarely follow, and shows the figures as they are on the Radar page.
An optical image and a radar image are fetched for seven sites near Finland. They are compared with images from a year earlier, and a weekly change is also calculated from the radar images. At ten-metre resolution you can make out construction, cleared ground and large equipment. The images are interpreted by hand every week.
Every evening HÄIVE lists the vessels moving slowly near the centre line of six subsea cable and gas pipeline corridors in the Gulf of Finland. For each vessel you see the name, flag, speed and the number of evenings in a row it has been on the list.
A satellite's thermal sensor picks up hot spots on the ground. HÄIVE counts the day's detections around 19 sites: garrisons, air bases, ports and refineries.
For four areas HÄIVE calculates the share of aircraft that reported degraded positioning during the day. The figure is compared with the area's own 14-day level.
Each measurement also states what it does not prove. A slow vessel near a cable is usually ordinary traffic, and a thermal detection may be a refinery flare. A measurement tells you what the sensor detected. The interpretation is made separately and written down where you can see it.
The model has 64 indicators. Of these, 34 measure threat and 30 measure the resilience of society. Four states and their criteria have been written down in advance for every indicator, and an observation changes the assessment only if it meets a criterion.
There are 47 sources, and each is rated on the NATO scale in two ways: how reliable the source is and how credible its information is. Rhetoric or a single anonymous source does not move the model.
Capability and intent are assessed separately. Building capability is expensive and slow, whereas intent can change quickly, so they appear in different indicators. That is why the strategic threat can be elevated while the conventional threat is low.
Automation collects and checks the material, and a person makes the interpretation.
Much of the security debate is interpretation written from inside one narrative or another. The same event looks like a threat or like exaggeration, depending on who is telling it.
The Discussion page goes through the narratives spreading on social media. For each one it sets out what is claimed, what is known, why the narrative spreads and which reasoning error it contains. How loud a narrative is has no effect on the model's figures, because otherwise the model would start measuring what people talk about.
The same discipline applies to HÄIVE's own assessments. Every interpretation is recorded with a competing explanation and the observation that would settle which one is right. The weekly Big picture and Signal assessment show the assumptions the assessment rests on.
A forecast is recorded with its date, deadline, probability and resolution criterion. When the deadline expires, the forecast is scored with a Brier score, where zero is a perfect forecast and 0.25 is the level of a coin toss.
Customers can read the whole forecast ledger, including the forecasts that missed. Accuracy is always calculated the same way, so you can follow how it develops yourself.
A drone found in Pellinki was confirmed as Russian-made three days after the forecast. It had come down before the forecast was recorded, but its origin was not yet known. The entry stays in the ledger as it is.
An analyst or consultant gets the groundwork for a situational picture ready-made: dated observations, sources and criteria in one place. The time that used to go into searching is left for interpreting the client's situation.
Every change of state comes with a date, a criterion and sources, so it can go straight into the risk register and the management team memo.
The Big picture, the preparedness level and the weekly review tell you whether something has changed enough to call for decisions. The grounds for the assessment can be put to the board as they stand.
The ten theatres of the Outer ring cover, among other things, alliance cohesion, Russia's internal resilience, sea routes and energy. They show which international change is on its way to Finland and to supply chains.
The power system, sea transport, Baltic Sea infrastructure and the energy markets each have their own indicators in the model, and traffic in the subsea cable corridors is shown in Radar every evening.
The Discussion assessments and the level-by-level guidance are a ready basis for a staff notice, also when a rumour is spreading fast.
The criteria are designed with you, the observations are tied to sources and the state updates automatically. The results are visible in HÄIVE only to your users, and your data is encrypted.
HÄIVE measures Finland's resilience through six lenses. The same model is built for your organisation, and it shows where a disruption would stop and where it would get through.
Do you notice a disruption in time?
Is it clear who is responsible for it?
How quickly do operations return to normal?
Does it show from outside that you are a hard target?
Do your people stay with you when the situation tightens?
How long can operations last in an emergency?
The method suits developments that show signs in advance: delivery times, raw material prices, market turns, competitors' moves and changes in regulation.
A fictional example.
The scope and schedule of an assignment are agreed separately. Design usually takes two to four weeks.
Ilkka Rajala began developing the system in 2025 for his own use, to make the news flow easier to follow and to see the systemic structure behind events.
The system proved its worth quickly and grew into HÄIVE, which was tested extensively before it became a service.
For 20 years Ilkka has coached people to observe and decide under pressure, among them business executives, elite athletes and special forces soldiers. For organisations he has carried out systemic analyses that read both language and behaviour.
From 2015 to 2020 he developed AI that identifies from speech and text how a person's thinking is structured. Ilkka used the method in corporate mergers, and it was also developed for intelligence purposes.
In HÄIVE he applies the same expertise to the security situation: how to tell a weak signal from noise without letting the analyst's own bias decide the result.
From half an hour to an hour. We go through what is worth following in your operations and who would use the service.
For one month you have the whole service, Tracking included, and you receive four weekly reviews. At the end we assess together how much value HÄIVE added.
The price of an annual contract depends on the number of users, the content you choose and how the material is used. You receive a written proposal with prices.
During the contract you can ask Ilkka Rajala about the assessments directly.
The trial fee is credited against the annual contract.
HÄIVE is a situational picture service that follows Finland's security situation, above all the military threat. News, 64 indicators of threat and resilience, its own measurements and a weekly review are in one encrypted web service used in a browser.
The news shows what happened. HÄIVE also shows whether the event changed the assessment. Every observation is compared with criteria written down before the observation came in, and a change of state leaves a date and a source. Most news changes nothing, and you can see that too.
Management and the board are asked from time to time whether something has changed enough to call for a decision on an investment, a supply chain, staff instructions or the continuity plan. With the picture already put together, the answer can be justified and entered in the risk register with its sources.
The Outer ring page follows ten international theatres, among them alliance cohesion, Russia's internal resilience, sea routes and energy. Each theatre has a state of its own, and the page shows the route by which a change would affect Finland and supply chains.
Yes. Hybrid action is one of the six threat lenses, and critical infrastructure and information influence and cyber operations have lenses of their own. The Radar page also shows vessels moving slowly in subsea cable and gas pipeline corridors and satellite navigation interference, which HÄIVE measures itself every evening.
OSINT is open-source intelligence: all the material is publicly available. It includes official bulletins, news, research institute reports, and satellite and measurement data. There are 47 sources, each rated on the NATO scale for the reliability of the source and the credibility of the information. HÄIVE is not an assessment by the Finnish authorities, the Finnish Defence Forces or NATO.
Every evening the model calculates a preparedness level on a five-step scale. Guidance has been written for each level on what to decide and do, and it can be used as a basis for staff instructions. As an assignment, the same six-lens resilience model is built at the scale of your organisation.
Automation collects and checks the material, and a person makes the interpretation. Forecasts are recorded with a probability, a deadline and a resolution criterion, and expired forecasts are scored with a Brier score. Customers see the whole forecast ledger, including the forecasts that missed.