Open data · Aggregation · Analytics · AI · Riga, Latvia

Scattered data, made usable.

Datti collects, cleans and connects data — public registers, open datasets and your own systems — then builds the analytics and tooling that make it worth having. Small team, direct contact, no layers in between.

[email protected]

Company: SIA Datti
Focus: Open & aggregated data
Also: Custom IT services
Contact: [email protected]

What we do

Five things, done properly. Most engagements combine two or three of them.

Data analytics & reporting

Recurring reports and one-off deep dives: the numbers behind a decision, explained in plain language instead of handed over as a raw export.

Reporting
Dashboards
Ad-hoc analysis

Open data aggregation

We locate, harvest and normalise public registers and open datasets, reconcile them against each other, and keep the result current.

Harvesting
Normalisation
Monitoring

Data engineering & pipelines

Ingestion, transformation and storage that runs on schedule, validates itself and tells you when something upstream breaks.

ETL / ELT
Scheduling
Data quality

APIs & integrations

Clean, documented interfaces over messy sources, so your systems exchange data automatically instead of through manual files.

REST APIs
Integrations
Automation

Custom IT services & help

Internal tools, small web applications, and a second opinion when the stack needs a decision — practical IT help without an agency layer.

Custom tools
Consulting
Support

AI on your own data

A model is only as good as the data underneath it. We build that foundation first, then apply machine learning where it measurably helps — and say so plainly when it does not.

Document and text extraction

Contracts, invoices, registers and PDFs turned into structured, queryable records using language models, with validation so that extraction errors surface instead of quietly spreading downstream.

LLM
OCR
Structured output

Retrieval over your own documents

Question answering grounded in your files and databases, with every answer citing the source it came from, so the result can be checked rather than taken on trust.

RAG
Embeddings
Semantic search

Classification and entity resolution

Categorising, matching and deduplicating records at a scale where hand-written rules stop working — reconciling the same company, person or address across registers that disagree.

Classification
Matching
Deduplication

Forecasting and anomaly detection

Volumes, demand and risk signals modelled on your own history, with alerts when the incoming data stops behaving the way it used to.

Forecasting
Anomaly detection
Monitoring

AI readiness review

An honest assessment of whether your data can actually support the model you have in mind, what is missing, and what it would cost to close the gap — before anything is built.

Consulting
Data readiness
Scoping

We do not sell AI as an end in itself. If a query, a rule or a spreadsheet answers your question reliably, we will tell you that instead.

How we work

Short cycles, visible output early, and documentation you keep whether or not we stay involved.

Scope
One conversation to establish the question worth answering and the data that actually exists.
Aggregate
Sources collected, cleaned and joined, with the gaps and caveats written down rather than hidden.
Build
Pipeline, API, report or tool — whichever is the smallest thing that answers the question reliably.
Hand over
Documented, automated and yours. We stay for maintenance only if that is what you want.

An example of the segment

Own platform

atvertiedati.lv

atvertiedati.lv is a platform we develop and maintain ourselves, bringing Latvian open data together in one place where it can actually be searched, compared and reused.

Shown here as an example of the segment we work in — not as a product offer.

Visit atvertiedati.lv

Domain: Latvian open data
Role: Built & maintained in-house
Status: Live, actively developed
Stack: Aggregation + web platform

Get in touch

Tell us what data you are stuck with, or what you need built. A short description is enough to start — we reply with an honest read on whether we are the right fit.