Give your customers a data scientist at their fingertips.

Locusive Analyze allows your customers to avoid clicking around your dashboard endlessly for the insights they need and instead just ask. Our powerful text-to-SQL agent is embedded right within your product, and gives answers in seconds.

Jamie Domini (Scented Candles)
What are our most popular products sold by unit over the past year?
Locusive
Generating SQL query...

Most reporting & analytics dashboards miss the mark

CSMs waste hours pulling metrics
Your reporting dashboards usually don't have all the data your customers need, so they just ask their CSM to compile and calculate the metrics manually.
SQL as a "gatekeeper"
For any reporting needs that fall outside the scope of your dashboard, an engineer needs to write a SQL report which disrupts their work and frustrates impatient clients.
Dashboards can be too complex or too limiting
Even when you have dashboards set up, they aren't always useful because they are configured to measure one thing and don't allow you to explore the data.
Interpreting and analyzing data can be complex
Interpreting and analyzing data from multiple sources in multiple formats can be time-consuming and frustrating. More often than not, it simply isn't done.

Replace complex database queries with plain English.

With the Analyze skill, customers get the answers they need from Locusive simply by asking. Locusive can interact with customers directly on your product or via Slack.

SELECT image, name, AVG(purchase_price)
FROM users
WHERE order_count > 1
What's the average purchase price for repeat customers?

Low code skill templates that are ready to go.

Great insights are only possible if you've built the reporting your customers need. With Locusive, you don't need to configure reports or write countless SQL queries. Simply create a few basic training templates to ensure Locusive can give customers they insights they need, from the right place, in seconds.

Data alerts that require zero technical configuration.

In addition to performing advanced data analysis, Locusive Analyze allows customers to set up alerts for specific data, on whatever frequency they want, simply by asking.

The security you need, without the hassle.

Locusive follows the user-based permissions you have setup, rather than forcing you to create policies. You can rest easy knowing that customers only have access to the data that is pertinent to them, not to other customers' data.

See the reasoning and raw data that led to a response

When you get a response from Locusive, you'll be able to audit how the system came to its conclusion, while also exporting the raw data to a CSV in case you want to run the analysis yourself.

A teachable system that understands your data

Locusive can query your database as soon as it's connected, but we also give you the tools to make it as accurate as possible

Include the tables and columns you need, leave the rest out.

Ensure customers are only accessing the data you want them to see, instead of a free-for-all.

Descriptions of your schema to help improve accuracy

Locusive injects details about your tables and columns with every database query to optimize the accuracy of your results

Reference guides to provide business-specific context and terms

Reference guides let you provide details about your database structure, how data should be joined, and any other context that's needed to create accurate queries.

Example queries to help the system learn how to pull the right data

Use examples to help the system understand how to run the right query at the right time, and templates with substitutable placeholders to ensure the system runs exactly what you need.

Learn More

Common questions on how we work with databases

Here's what folks ask us when they hear about our text-to-SQL and database query capabilities...
How do you ingest our data?

Nobi connects into a database system that you already have running. This way, there's no database snapshots, data transfer, or complicated ETL jobs that anyone needs to worry about. You give us a read-only user (usually to a read-only replica of your main database), and our system automatically detects your schema and runs queries on that database.

How do you securely store credentials?

When a new organization signs up to Nobi, we create a secret encryption key for them that's stored in our secrets manager. That key is then used to encrypt their database credentials, which are stored in a separate system, and are used to connect to the database whenever the copilot needs to run a query.

How does your system figure out what it needs to do?

When one of your users sends in a question, our copilot previews all of the data sources it has available to it and makes an informed decision about which system or systems it needs to query to get the answers to your user's question. It then queries those systems to get the data it needs, retrying or falling back to other systems whenever it needs to.

Do you support my database?

We support some of the most popular databases and data warehouses in use today. You can see more here.

How do I teach your system how to make accurate queries?

We give you several different ways to provide enough context our copilot so that it can query your database properly. We recommend creating a Reference Guide that provides high-level and nuanced details about your database to our system, creating several different example queries that the system should run to answer different questions, creating templates that the system should run exactly when specific questions come in, creating training instructions to tell the system how to behave when specific types of requests come in, and including the tables and columns that you need the system to use when users send in a request.

Does your system only generate the SQL or do you execute queries as well?

Our system executes queries and can even summarize and analyze the results for you.

I have more questions. Where can I find answers?

We'd love to help. You can contact us here on our website, or set up time to walk through your own needs on a demo call.