Smriti for AI-native analytics

Modern data stack for AI-native analytics

Ask in plain English. Smriti turns your business context into SQL, runs it against your data, and shows the answer your team can trace.

We will use this email to follow up about Smriti.

Smriti
Which region missed target last month?
South missed target by 8.4 percent. The gap came from distributor returns and slower repeat orders.
SQLSELECT region, target_gap_pct FROM sales_monthly WHERE month = 'May' ORDER BY target_gap_pct ASC;

Ask the question you would normally send to an analyst.

Sales, finance, and operations teams get answers from approved data without waiting for a dashboard change or a SQL pull.

Ask questions in plain English.
Compare periods, regions, products, customers, and teams.
Ask follow-ups in the same thread.
Get answers from your approved business data, in one place.

Start with the questions already in the queue.

Which region missed target last month, and by how much?
Who has not paid in 60 days?
Did the festive scheme lift secondary sales?
Which products are slowing down in the South region?

Smriti remembers how your business talks.

Teach Smriti your tables, joins, metrics, and company vocabulary once. After that, your team can ask questions in the language they already use.

01

Teach the business context

Define terms like net sales, active customer, region, route, or net 30 in one shared layer.

02

Ask in plain English

A user asks the question they already have, without choosing a dashboard or writing SQL.

03

Run readable SQL

Smriti writes a query, runs it against your data, and shows the SQL behind the answer.

04

Keep the conversation going

If the answer needs more detail, ask a follow-up. If the question is unclear, Smriti asks before it runs.

Answers your team can inspect.

Smriti does not invent the number. The answer comes from a query. The SQL, result table, and chart stay together so business users move faster and data teams can audit the work.

SQL-backed answers
Business vocabulary
Clarifying questions
Answers you can audit

Start with the data you have.

Available now

  • Natural-language questions
  • SQL-backed answers
  • Business terms and semantic definitions
  • Workspace-level access
  • CSV and Excel upload

Coming

  • Direct ERP connectors
  • Direct database and warehouse connectors
  • Per-person access controls

Do the numbers come from the AI model?

No. Smriti writes SQL, runs it against your data, and returns the result. You can inspect the query behind the answer.

Do we need a data warehouse before we try it?

No. You can start with CSV or Excel data. Direct connectors are coming.

Can the data team control definitions?

Yes. The data team defines the terms, joins, and metrics Smriti uses.

Who should use Smriti first?

Start with a team that asks repeat data questions every week: sales, finance, operations, or a data team that handles the request queue.