Build text-to-SQL right into your product

Enable conversational data analytics, simplify business intelligence, and build database LLM-agents with the world’s first SQL API built with generative AI.

Used by teams at:

waii query create “Show me monthly sales per store that deviated more than 10% from average” WITH avg_monthly_sales AS ( SELECT AVG(ss_sales_price - ss_ext_discount_amt) AS avg_sales, d_month_seq FROM tweakit_playground.retail_data.store_sales AS ss INNER JOIN tweakit_playground.retail_data.date_dim AS dd ON ss.ss_sold_date_sk = dd.d_date_sk WHERE EXTRACT(YEAR FROM dd.d_date) = 2023 GROUP BY d_month_seq ) monthly_sales AS ( SELECT s.s_store_sk, s.s_company_id, i.i_brand, i.i_category, dd.d_month_seq, SUM(ss_sales_price - ss_ext_discount_amt) AS sales FROM tweakit_playground.retail_data.store_sales AS ss INNER JOIN tweakit_playground.retail_data.date_dim AS dd ON ss.ss_sold_date_sk = dd.d_date_sk INNER JOIN tweakit_playground.retail_data.item AS i ON ss.ss_item_sk = i.i_item_sk INNER JOIN AS s ON ss.ss_store_sk = s.s_store_sk WHERE EXTRACT(YEAR FROM dd.d_date) = 2023 GROUP BY s.s_store_sk, s.s_company_id, i.i_brand, i.i_category, dd.d_month_seq ) SELECT ms.s_store_sk, ms.s_company_id, ms.i_brand, ms.i_category, ms.d_month_seq, ms.sales, ms.sales - LAG(ms.sales) OVER (PARTITION BY ms.s_store_sk, ms.s_company_id ORDER BY ms.d_month_seq) AS prev_deviation, ms.sales - LEAD(ms.sales) OVER (PARTITION BY ms.s_store_sk, ms.s_company_id ORDER BY ms.d_month_seq) AS next_deviation FROM monthly_sales AS ms INNER JOIN avg_monthly_sales AS ams ON ms.d_month_seq = ams.d_month_seq WHERE ABS(ms.sales - ams.avg_sales) / NULLIF(ams.avg_sales, 0) > 0.1 ORDER BY ms.s_store_sk, ms.s_company_id, ms.d_month_seq


Get the most complete and accurate translation of plain English to SQL available. Waii combines a purpose-built compiler and optimizer with the latest AI models and the most complete coverage of SQL to achieve 90%+ accuracy compared to an industry standard of 60%.



Generate queries optimized for your database. No outages due to inefficient SQL. Push down the right operations to reduce data early, limit the scan set, and optionally use single pass algorithms.

cat “web-vs-store-spend-report.sql” | time waii query run Running... 1117.1 seconds waii query create “Find customers whose spending increase was larger for the web than in stores last year” | time waii query run let’s see how the generated query performs Running... 223.5 seconds
waii context add "An active user is someone who logged in at least once in the last 30 days." --schema web_traffic Context added. waii query create "Give me monthly active users in the last 2 years" SELECT date_trunc('month', login_timestamp) AS month, count(distinct user_id) AS active_users FROM web_traffic.user_activity WHERE login_timestamp >= now() - interval '2 years' AND login_timestamp > now() - interval '30 days' GROUP BY month ORDER BY month ASC;


Produce SQL that’s 100% in tune with your business. Automatically build a persistent semantic layer to customize the system to your business and database, not the other way round.

Used by teams at

What they say

Waii's text-to-SQL is showing impressive capabilities around dealing with complex relationships in our database.

Director of Data Services at Rev


No DB is too complex

The Waii API scales to thousands of databases, schemas, tables and columns. Proven to work with any layout: Snowflake/star schemas, operational and normalized databases and more.

Database icon
Handles real-world SQL

From complex join graphs, multi aggregate & window functions, UDFs, JSON, semi-structured data, geo-location, table functions, DML, DDL, streaming extensions and more.

Feature icon
Works out of the box

Hook up your database and start generating. Knowledge graph generation is automated; A built-in compiler/optimizer ensures perfect queries on: Snowflake, PostgreSQL, MySQL, Trino, ClickHouse, Athena, SingeStore, and more

Security icon
With proven security

Manage access to your database and models with fine grained controls. Limit permissions to metadata only. Protect sensitive data, such as PII. Encrypt at rest and in motion. Support federated identity and single-sign on.

Feature icon
Adapts to your needs

SaaS or self-hosted, SSO, public or private LLMs, flexible vector store choice, out-of-the-box data catalog & semantic layer integrations, wide range of DBMS and BI tool support. Waii fits into your architectures with no compromises.

Feature icon
And much more...

Go beyond query generation with full chat APIs, charts, interactive discovery and debug, semantic search, dialect translation, optimization, ...

Leadership Team

Gunther Hagleitner


SVP of Product and Engineering at Cribl, CVP of Engineering at Cloudera and Hortonworks, Engineering at IBM, Yahoo, and Amazon.

Wangda Tan


Engineering leader of SQL team at Snowflake, Engineering leader at Cloudera and Hortonworks, ML engineering at Alibaba Cloud.

Derek Chang

Head of Sales

Enterprise Sales at Snowflake, Verkada, and Cribl (3x unicorn)

Backed by First Round Capital and top angels


Can I deploy Waii in my environment?
What does it take to trial Waii?
What sets Waii apart from other text-2-sql vendors?
How does Waii control access for each user?
How does Waii handle PII data?
What models do you support and recommend?
What data does Waii share with models?
Does it take a lot of data modeling and cataloging to use Waii?
What databases does Waii support?
What data catalogs and semantic layers does Waii support?
Do I need a dev team to deploy Waii?
What vector databases does Waii support?
What is conversational intelligence and how does it relate to Waii?
Does Waii need a snowflake or star schema to use Waii?
Can Waii handle OLTP schemas?
What are Waii's scale limits?
Will you train your model based on any of my customer data?
Does Waii write optimized queries?

Get started with Waii today!

For an enterprise-level implementation, we offer competitive consumption-based pricing, billed either monthly or annually.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.