Historically Better

Hi! While we are keeping on working on our Semantic Layer, with very promising results, we have also made some very nice quality of life improvements for Wobby Explorer and Agent History.

Agent History is gone, Agent Monitoring is in

We have completely revamped the Agent History page for easier navigation, helping data teams debug conversations easier.

  • Filters
    Easily filter conversations by follow-ups, channel, mode, feedback, and other indicators.

  • All info in one place
    See all relevant conversation metadata at a glance.

  • Star to keep
    Star your most relevant conversations to keep them handy.

Explorer niceties

While the improvements to explorer are very minor, heavy Explorer users will definitely appreciate them:

  • Star and delete
    Star the important chats to keep them at the top, and delete the clutter to have a tidy workspace

  • Rename chats
    We know auto-generated names are not always ideal, so you can now rename chats

  • Debug quickly
    If you are a data person, you can click "Debug in Studio" button in Explorer to quickly open the same conversation in Studio. (Especially useful when opening conversations from Slack!)

Bug fixes

  • Wobby was a bit overworked
    We have done a lot of work to decrease the latency of requests, so Wobby should respond much faster. If you are still experiencing problems, ping us and we will look into it.

  • Charts not charting
    Sometimes, the charts in Deep Analysis reports would just not show. Now they should.

Thanks for reading, and have a great week!

Query Caching

Honestly, this is about the most boring feature in our entire product, but if you use SQL templates, and your data source is a little bit slow, it makes all the difference, so we decided it deserves its own changelog entry to show that even the most boring things can make the biggest difference.

The feature is opt-in, so if you have strict data handling requirements, you can keep it disabled to maintain absolute zero copy data handling. However, if you don't have any special data handling policies, go ahead and enable it by going to the data source settings, and clicking "Cache Settings" in the corner.

Housekeeping

Hey! We are currently quite busy building the next big thing for Wobby, and if you were at the Data Expo in Utrecht you already had a sneak peek at our CEO's talk. In the meantime, while we prepare to unveil the big news, we did some housekeeping:

Spring Cleaning for Your Data

Remember that tedious task of manually cleaning up deleted columns from your schema? Neither do we - because it's now automatic! When columns vanish from your database, Wobby tidies up after them like a well-trained data butler. No ghost columns haunting your schemas anymore.

Agents Listen To Instructions Better

Our agents went back to school and learned to actually read the instructions. They now inspect specific columns and tables with the precision you'd expect from a data analyst who's had their morning coffee. Your detailed guidance about which data to analyze no longer falls on deaf ears.

Database Expansion Pack

  • Redshift Goes Multi-Schema
    Why settle for one schema when you can have them all? Connect to multiple schemas in a single Redshift instance and let your agents roam free across your data landscape.

  • PostgreSQL Materialized Views
    Because sometimes regular views just aren't fast enough. Your agents can now tap into materialized views for that sweet, sweet pre-computed performance boost.

Interface Polish

  • Loading Spinners That Know When to Stop
    Fixed those eternally spinning loading indicators that made you question if the task was done or if you'd entered a temporal loop

  • Unified Studio Messaging
    The input box in Studio got a makeover - controls now play nicely together in one cohesive interface

  • Data Source Stability
    Removing your last table from a data source no longer triggers the nuclear option of deleting everything. Because sometimes you just want to start fresh, not burn it all down.

Databricks, MS Fabric, Reply & Explain

New Features

Fresh Integrations

  • Databricks is here!
    Connect your Databricks data warehouses and bring the power of Wobby to your entire lakehouse architecture. Your data engineers will love this one.

  • Microsoft Fabric Support
    We've heard you, Microsoft shops! Full integration with Microsoft Fabric is now live, complete with Azure AD authentication. One less barrier between your teams and their data insights.

Explorer Gets Smarter

  • Reply & Explain
    Ever wondered how Wobby got to that answer? Now you can simply highlight any result and click "Explain" for the full breakdown. Want to dig deeper? Hit "Reply" to continue the conversation. It's like having a data analyst explain their work, step by step.

  • Deep Analysis Follow-ups
    Those comprehensive Deep Analysis reports you love? They're now conversational! Ask follow-up questions to clarify findings or explore new angles without starting over.

Major Improvements

Better AI, Better Analysis

  • Actual Report Summaries
    Deep analysis reports now start with meaningful, AI-generated summaries that actually tell you what was found (goodbye, generic placeholders!).

  • More Visualization Options
    We've added stacked column and bar charts because sometimes you need to see how parts make up the whole. Your presentations just got prettier.

Data Management Made Easier

  • CSV Gets Some Love
    Metadata quality scores aren't just for fancy databases anymore – your CSV files now get the same treatment.

  • Smarter Schema Syncing
    Changed your database schema? We've got you covered. Wobby now properly syncs changes, including automatically removing columns you've deleted. No more ghost columns haunting your analyses!

Bug Fixes

Database & Query Improvements

  • BigQuery Playing Nice
    Foreign keys in BigQuery? We see them now. Those timestamps that looked like JSON gibberish? Fixed – they're readable dates again.

  • SQL Templates Working Everywhere
    Parameter binding issues with Microsoft Fabric and MSSQL? Sorted. Your templates work consistently across all connections now.

  • MySQL Stability
    Squashed that pesky edge case causing connection timeouts. Your MySQL queries should run smoothly now.

Explorer & UI Polish

  • Smoother Task Submission
    No more endless loading screens when submitting tasks. Click, submit, analyze – as it should be.

  • Charts That Chart
    Fixed various rendering issues so your visualizations actually... visualize.

  • Clearer Login Experience
    Made the email verification and 2FA screens clearer because nobody should struggle just to log in.

  • Better Error Messages
    When data source connections fail, you'll now get helpful error messages instead of cryptic codes.

Performance & Reliability

Behind-the-Scenes Improvements

  • Reports That Finish
    Deep report generation no longer mysteriously stops halfway through. Your comprehensive analyses will complete as expected.

  • Charts That Work
    Fixed composer failures that were preventing certain chart types from generating properly.

  • Data Display Logic
    Resolved an issue where data rows were showing up when they shouldn't. Less confusion, more clarity.

Communication Enhancements

  • Slack Integration Stability
    Cleaned up several Slack issues including those annoying duplicate Explain buttons and improved response handling. Your Slack-based analytics should feel much smoother.

Questions or feedback? We'd love to hear from you! Reach out to your account manager or drop us a line at support@wobby.ai

MotherDuck, Passkeys, and a new lick of paint

What's New

Fresh Data Connections

  • MotherDuck is here!
    We've added support for MotherDuck cloud data warehouse—because your data should fly, not waddle

  • Smoother Data Management
    We've made selecting tables in your agents way more intuitive, plus you can now import and export metadata with just a few clicks

Security That Just Works

  • Passkeys Have Arrived
    Sign in with just a touch—no more password gymnastics required

Explorer Gets Smarter

  • Deep Analysis Leveled Up
    We've made it easier to follow along as your agent tackles complex analysis tasks with better progress tracking and clearer step visualization

Your Agents, Now Even Sharper

  • SQL That Actually Makes Sense
    Our query generation got a serious upgrade—expect more accurate templates that actually understand your data

  • Knowledge → Semantic Layer
    We've renamed Knowledge to Semantic Layer (because that's what it really is!) and added bulk operations plus those handy verification badges you've been asking for

  • Quick Debug Mode
    Spotted something odd? Jump straight from Explorer to Studio to see what's happening under the hood

Things We Made Better

A Cleaner, Clearer Interface

  • Explorer Got a Makeover
    We've redesigned the chat interface with better spacing and a layout that actually makes sense

  • Data Sources That Don't Confuse
    Reorganized navigation with clearer naming and consistent layouts throughout

  • Charts That Actually Chart
    Fixed those pesky horizontal bar chart labels and modernized our entire charting engine

Speed & Stability Wins

  • Explorer Loads Faster
    Home page loads quicker, less waiting around for your data insights

  • Errors That Help
    When something goes wrong, you'll actually understand what happened

  • Scrolling That Follows Along
    Chat conversations now scroll properly—no more losing track of the conversation

Squashed Bugs

The Big Ones We Fixed

  • Apparently, our authentication worked wonders except for cases when it mistakenly locked people out—no more of that now.

  • Charts show up instead of mysterious blank spaces

  • SQL queries execute without throwing tantrums

  • Organization names display correctly without breaking the entire layout

  • Mobile users can now actually complete onboarding in Explorer

Explorer & Studio Polish

  • That share button? It shares things now

  • Table previews stay in their lanes

  • Logo clicks take you where you'd expect

  • Deep Analysis comments have room to breathe

  • Progress trackers in Deep Analysis actually track progress

Under the Hood

  • Queries Run Smoother
    Database queries are more reliable and noticeably faster

  • Tasks That Don't Give Up
    Better retry logic means fewer failed tasks interrupting your flow

Better Integrations

Slack & Teams Love

  • Slack charts now share beautifully (no more screenshot workarounds!)

  • Messages format correctly, attachments attach properly

Data Platform Harmony

  • Setting up data connections is less of an adventure, more of a breeze

  • Metadata import/export now properly handles primary keys

  • Schema and table selection works consistently across all platforms