From raw ore to insight
Import your scattered datasets, trigger SMS, RCS and enrichment, and ask your data questions in natural language. From the raw file to a measured send, in one place.
A file comes in and becomes a searchable index, with sensitive fields protected.
Providers enrich the records and send by SMS and RCS; the result flows back to each record.
You ask in natural language and the answer arrives formatted, with a chart.
Charts go to a dashboard that updates itself.
From the raw file to insight, in one platform.
The problem
Customers in a spreadsheet, the portfolio in another system and this month's export in a shared folder.
The campaign goes out from the provider's panel and what came back never reaches the customer record.
Every cut of the data depends on someone writing the query or building the report.
How it works
Real app screens, with fictional data.
CSV, XLSX, JSON or NDJSON become a searchable index, with the schema and types detected. CPF, CNPJ, phone and email are identified automatically as sensitive and protected in queries and charts.
A flow filters the data and triggers the provider: data enrichment, SMS or RCS. Each run shows progress, sent messages, failures and the amount debited from your balance.
Ask in natural language, across one or more indexes. The answer arrives in Markdown, with a chart you can switch to another type, export as CSV or PNG and send to a dashboard.
Global filters apply to every widget, KPIs get an icon and the dashboard updates in real time.
Ask your data
This is how the Data Assistant answers: formatted text and a chart ready for the dashboard.
Sample questions
Which states hold the most customers in the Retail Customers index?
SP leads by a wide margin, followed by MG and RJ. Together the three account for more than half of the dataset; the remaining states form the Pareto tail.
How are customers distributed by segment?
Bronze is the largest segment, followed by Silver. Gold and Platinum are smaller and together form the highest-value group in the dataset.
Compare the Retail Customers and Collections Portfolio indexes by state.
I queried both indexes and cross-referenced the totals by state. SP has far more records in Retail than in Collections; in MG and RJ the two datasets are close.
Fictional data. In the app, the answer comes from your indexes.
Featured capabilities
Ask in natural language. The answer comes formatted, with a chart.
The Data Assistant queries your indexes, compares more than one in the same conversation and cross-references the analyses. Conversations can be renamed and deleted. OpenAI or Azure OpenAI models.
See the feature: AI on your dataChart types
SMS, RCS and enrichment in one place.
Each provider has its own return URL. The platform administrator can register other providers that expose an HTTP API.
See the feature: Providers and channels
Every dataset becomes a searchable index.
Paste a sample document, check the inferred types and mark the key field.
See the feature: Data indexing
The answer becomes a widget.
KPI with an icon, Pareto, donut and table, with global filters and real-time updates.
See the feature: DashboardsPlatform
Filters, provider and recurrence. The run goes out at the scheduled time.
A per-user API key for the integration API and one return URL per provider.
Pay-per-use debits, history and consumption visible in real time.
One partner's data never shows up for another.
CPF, CNPJ, phone and email are identified automatically and protected in queries and charts.
Who it is for
The pain: Data in spreadsheets and exports; campaigns sent without knowing what came back.
Segment your data with one question and see the result of every send on the record.
What matters most
The pain: A huge portfolio, priorities set by gut feeling and SMS results lost along the way.
Know who to contact first. The portfolio Pareto comes from one question.
What matters most
The pain: Every client has a different provider and a different return format.
Connect the client's provider through API and webhook. The result lands on the right record.
What matters most
The pain: A queue of report requests and manual cross-referencing of datasets.
Cross portfolios without writing SQL and publish a dashboard that updates itself.
What matters most
Quick comparison
Side by side, with what each option really solves.
| Capability | DeepMiner | Spreadsheet + BI tool | Standalone sender |
|---|---|---|---|
| Import CSV, XLSX, JSON and NDJSON into a searchable dataset | Yes | Partial Opens the file; joining datasets is manual | Partial Only the send list |
| Ask in natural language and get a chart | Yes | Partial Formula, query or a prebuilt report | No |
| Send SMS and RCS from a filter on your data | Yes | No | Yes |
| Result of every send written to the customer record | Yes | No Manual cross-referencing | Partial Report in the sender's panel |
| Dashboards with global filters | Yes | Yes | Partial Send reports |
| Scheduled flows on top of your data | Yes | No | Partial Schedules the send of a list |
Frequently asked questions
CSV, XLSX, JSON and NDJSON, plus JSON records sent through the API. Each dataset becomes a searchable index with detected schema and types.
SMS and RCS sending and data enrichment come ready to use. The platform administrator can register other providers that expose an HTTP API, with method, payload template, batch sending and webhook return paths.
Each provider has its own webhook URL. Events arrive through it, find the matching record and update its status, with the history of the latest events kept on each record.
The Data Assistant uses OpenAI or Azure OpenAI models. It queries the indexes you choose, cross-references analyses from more than one index and answers in Markdown, with a chart.
No. The platform is multi-tenant: one partner's data never appears for another, and sensitive fields such as CPF, CNPJ, phone and email are identified automatically and protected in queries.
Through a prepaid balance, debited per use. The balance, the history and the progress of each run show up in real time in the app.
Access starts with the demo: partners are created by the platform administrator. If you already have access, use the Log in button.
Book a demo and watch a dataset get indexed, enriched and queried.