The advanced search filters build a list of companies from five kinds of condition: sector, financial figures, registration date, related persons and status flags. The result downloads as a CSV file, and no account is needed.
Latvia has 3,282 active companies in the construction of residential and non-residential buildings — NACE division 41.1 With revenue above a million euro there are 383. Of those, 41 carry at least one status flag: an insolvency proceeding, a liquidation or a suspension of economic activity by VID.2 That leaves 342 to work with.
How the filters work
Each advanced filter is a condition added to the previous one, and the count in the results bar changes immediately.
| Filter | Companies |
|---|---|
| Active companies | 221,182 |
| + sector NACE 41 | 3,282 |
| + revenue from EUR 1,000,000 | 383 |
| − status flags | 342 |
The financial filters have five rows: revenue, profit, assets, equity and employees. Each has its own minimum and maximum, so a range can be bounded at both ends. The sector field takes a NACE section letter, a two-digit division, an exact class such as 41.20, or free text matched against the activity description. The registration date has a start and an end bound, so a single year's cohort takes one step.
Three basic selectors sit above the panel: legal form, status and region. Below it are seven quick searches that load a ready set of conditions. Riga, for example, has 99,654 active SIA.3
The filters live in the address, so a selection is a link: send it, bookmark it, open it a month later with the same conditions and today's data.
The filters that are not numbers
Two groups answer not "how large" but "who stands behind it".
The person filters find companies by the name of an official, a shareholder or a beneficial owner. Latvijas Republika in the shareholder field returns 42 companies.4 A name search does not produce that list: the term appears in none of the 42 names.
Status flags work the other way round: they show which companies to strike off. Across the register, 15,884 companies carry an insolvency record, 11,104 a liquidation, 10,720 a suspended activity, 37 a sanctions flag.2 In the construction example the flag falls on 41 of the 383 — 35 insolvency, nine suspension, two liquidation. Some carry more than one.
What the list is for
To a sales team it is a segment. The CSV file has fourteen columns, from registration number and address to revenue, profit, equity and employees. One file holds no more than a hundred rows.5 That is enough to sort by size or region before the first conversation.
To procurement and finance it is the check before a contract. In the construction list those 41 carry a record that appears in neither the name nor the website, and it is visible before the call rather than after the first unpaid invoice.
To an analyst it is the size of a sector: three thousand active companies, of which 383 clear a million in revenue, is a different picture from the sector total. To a journalist or a researcher the person filters are worth the most, because they show connections a single company page does not.
When filters run out
Not every question is a combination of fields. A comparison between years does not fit the filter bar, nor does sorting by a measure with no filter of its own, nor a condition that depends on another. Those go to the AI analyst in plain language, and the answer comes from the same database as the filter counts, not from a model's memory.
What follows the list
The list is a beginning, not a result. Every registration number leads to a company page with financial history, officials, shareholders, beneficial owners, the network of connections and a risk assessment. A company worth watching goes on a watchlist, which reports when its record changes. The same selection is available through the API.
That is the difference between a list and working from a list: the first can be bought as a file, the second needs a full record beneath every row.
The filters stay in the address, so the same selection can be repeated later. If a quarter from now fewer than 342 are left without a flag, what changed is the sector, not the filter.
Notes
Footnotes
-
NACE is the EU classification of economic activities. Division 41 covers the construction of residential and non-residential buildings. Izlūks data, retrieved 23 August 2026: 3,282 active companies carry a code in division 41; 383 of them report revenue of EUR 1,000,000 or more in their most recent filed annual statement. ↩
-
VID — Valsts ieņēmumu dienests, the State Revenue Service — may suspend a company's economic activity, for example over unfiled returns or unpaid tax. Register-wide figures on 23 August 2026: 15,884 companies with an insolvency record, 11,104 with a liquidation record, 10,720 with suspended economic activity, 37 with a sanctions flag. These cover the whole register, not only active companies. Excluding insolvency, liquidation and suspension from the 383 leaves 342. ↩ ↩2
-
SIA — sabiedrība ar ierobežotu atbildību — is the Latvian private limited company, the dominant legal form in the register. ↩
-
The shareholder filter matches the name exactly against shareholder records. The query
Latvijas Republikareturned 42 companies on 23 August 2026. Broader searching by person is what the separate persons tab is for. ↩ -
The CSV export returns up to 100 rows in one file, with the columns
regcode,name,type,type_text,registered,closed,address,activity,revenue,profit,total_assets,equity,employeesandpaid_up_equity. The file follows the same filters that are in the address. ↩
