Skip to content
Predict.ai
Data

Connect your data. Then explore it.

Bring in files, API events, streams, or warehouse data. Predict.ai organizes them into signals you can explore, question, transform, and use in forecasts.

One workspace — goals, signals, segments, and pipelines wired together.

Signals

Stream data in with one API call.

A signal is a named time series: a key plus timestamped values. Push one through the API, import a file, or connect a source; Predict.ai profiles it and makes it available for analysis.
  • Files, REST, Kafka, S3, BigQuery, Postgres, webhooks
  • Events, too — discrete occurrences, scored for relevance automatically
  • Analytics, the AI analyst, and forecasting use the same signal layer
Every signal in the workspace — freshness, trend, distribution, relevance.

Segments

Your data, aligned and ready for training.

Combine events and time series in one normalized grid. Clean, align, fill gaps, and engineer features once; downstream trainings reuse the same versioned definition.
  • Source-agnostic — segments don't care where data came from
  • Built-in cleaning, alignment, and gap-filling
  • Engineered features become first-class columns
Pick targets and features, choose a fill strategy — the grid stays clean.

Automatic data insights

Understand what changed—even without building a forecast.

After data lands, Predict.ai profiles distributions, anomalies, recurring rhythms, correlations, and lead-lag relationships automatically.
  • A concise brief highlights important changes
  • Anomalies, seasonality, and relationships in one panel
  • Supporting signal history stays visible for review
The intelligence brief — movers, leaders, rhythms, and anomalies, computed on land.

Multiple sources, mixed cadences

Files, REST, Kafka, S3, BigQuery, Postgres, webhooks — same workflow for every source.

Reusable data definitions

Models train against a versioned segment, so source changes can be reviewed and rebuilt consistently.

Lead-lag relationships

Compare which signals tend to move first and which ones follow.

Automatic profiling

Anomalies, seasonality, and correlations are surfaced after ingestion.

Keep existing sources

Connect databases, warehouses, object storage, streams, or the REST API.

Lineage end-to-end

Every cell in a segment can be traced back to its raw row, its connector, and its run.

Built-in connectors

  • Files & uploads

  • PostgreSQL

  • BigQuery

  • S3-compatible

  • Kafka

  • REST & webhooks

Try it on your data.

Connect an outcome and its history. Predict.ai finds the drivers, races the models, and keeps the forecast live — in the workspace, over the API, and through agents.

Start with your dataTalk to sales