DarkMath products
Three products on one foundation. DarkMatch resolves identity, DarkAgent turns it into decisions and supported actions, and DarkMirror finds more of the people who matter.
| Product | What it does | What you get | Status |
|---|---|---|---|
| DarkMatch | Resolves duplicate and fragmented records into people and households | Resolved IDs with confidence and sources, households, updated contact data, embeddings | Available |
| DarkAgent | Turns resolved data and market context into answers, recommendations and supported actions | Sourced answers, market comparisons, audiences, prepared orders, outcome reports | New |
| DarkMirror | Finds people who behave like your best customers, even with missing fields | Semantic audiences with counts, exclusions and contact coverage | Pilot |
How they fit together
The order matters. Identity comes first because every count, recommendation and audience inherits its errors. A customer split across four records looks like four weak prospects. Two people merged into one look like a customer who does impossible things.
DarkMatch fixes that foundation. DarkAgent reads it next to your market and recommends what to do. DarkMirror extends the result to people you don’t know yet. Outcomes flow back, so the next match, the next recommendation and the next audience start from better evidence.
Resolve with DarkMatch
Who is this, and which records belong together?
Decide and act with DarkAgent
What does it mean for us, and what should we do next?
Expand with DarkMirror
Who else looks like the customers we want more of?
Measure and feed back
Did it work, and what should change next time?
The foundation: the Cognitive Data Graph
A CDP stores records. The Cognitive Data Graph captures what they mean.
It connects resolved people, households, businesses, vehicles, observed and inferred attributes, events and outcomes. Every connection keeps its source, its time, its confidence and its permitted use, so an answer can always show where it came from and whether it may be used for the job at hand.
DarkMath converts each record into a vector embedding, a numerical description of its meaning, in a 768-dimension space. Records that mean the same thing sit close together even when their text differs. We call the pull that groups them Semantic Gravity: new data settles toward the identity it fits best, and each new record makes the next decision sharper.
Built to complement your stack
- Works alongside your CRM, DMS, CDP or warehouse. Nothing to rip out.
- File upload, SFTP, Amazon S3, Snowflake sharing or REST API
- CSV, JSON and Parquet accepted as is
- Vector-only ingestion when personal data should not leave your environment
- Hosted, in your environment, or dedicated clean room
Start with the product that fits your question.
Most teams begin with a DarkMatch test or a DarkAgent walkthrough on synthetic data.