You now have the ability to filter conversations by different things. This allows you to see the types of messages you want and exclude the ones you don't for better visibility!

Next week will bring a new wave of updates, but also there's been a lot of backend work done that you don't see to prepare for the new push of goodies... testing portal and adding HubSpot as a source. There are big changes coming that will make CloseBot even easier to use for all! 👏

You now have the ability to see when a follow-up is scheduled to be sent, but also to delete that follow-up or change the date and time that it will be sent from within CloseBot. This feature was prioritized to align with our focus on . We want you to have visibility into why and when things are happening at all times.

You'll notice a plethora of improvements on the user interface throughout the site. We've done a large push to go back through the entire app and clean up some things that have been put on the backburner... now that we are stable and finishing up the Conversations page udpates, we wanted to do some clean up for everyone 😊

Next week HubSpot will be in closed beta testing. There will also be some final work to the conversations page for log auto-opening and message filtering. Then the team turns focus to the testing portal of the Job Flow testing area to bring testing in-flow (showing you highlighted nodes as you test your agent).

CloseBot can now send images across all supported channels! Find this ability within the area of the Statement Action. A link to your hosted image is accepted and variables are accepted.

CloseBot now understands audio messages from all supported channels of our sources. On our end, audio is converted to text and we respond to this via text on the same channel. Nothing extra needs to be done on your end to enable this. UI is coming soon to display the audio messages, allow you to play them back, and show you our transcription. (Conversations tab doesn't currently display the text conversion or let you listen to the audio, but that's coming next week).

You're able to open a side contact card to see the values of fields updated by CloseBot for that contact.source.

Easily navigate to the contact within your connected CRM by clicking the link icon in the new contact card.

Hover over different icons within the conversation page to see what your Agent is doing, including details about the value of field updates and calendar availability pulled in from your calendar!

Webhook is now an option for a source within CloseBot. This allows you to integrate with any system that gives you control on outbound and inbound message triggers. Tie CloseBot into other non-native CRMs or your own solution! Check out the new help doc HERE to learn about the webhook source setup

Our API is incredible, allowing you to do everything you can do and view via the user interface, also programmatically. Learn more about our authentication for the API HERE. The resources here will continue to grow to support our fellow developer community! 👏

These display languages make it easier for our Portuguese, French and Spanish speaking users and clients to navigate the CloseBot site. We will continue to add additional languages. Remember, CloseBot agents already work for any language supported by your chosen AI Provider... this is purely for the navigation of the site itself.

The AI logs can be difficult to navigate. We have two improvements to this. Further improvements will continue to be rolled out to this page next week.

CloseBot now understands audio messages from all supported channels of our sources. On our end, audio is converted to text and we respond to this via text on the same channel. Nothing extra needs to be done on your end to enable this. UI is coming soon to display the audio messages, allow you to play them back, and show you our transcription.

Enabled by default, you can disable this if you want to make sure a custom scenario is only able to be triggered once.

Transaction information within your Wallet section will now have broken up values so you can see the reasons why your card was billed a certain amount for usage per category.

Slow load times due to our host have been partially resolved. We will continue to make updates with them to make sure our traffic volume is handled adequately across the globe. We've also made large improvements to the logic that affects booking accuracy. If you've had issues with booking in the past, go give it a try again as it's drastically improved! As always, please go to support first if you encounter issues that our team should know about.

We have planned out the roadmap of features needed to be complete before the HighLevel summit. Most of them are publicly listed on our roadmap, and some are private. There's a lot coming within the next 7 weeks and this week and next the team is laying a lot of the foundational pieces for those updates.

Take a look at our roadmap... we are hoping that 90% of the items listed as in-progress will be done next week, possibly along with some items from the category (and some items not on our roadmap at all 😉)

Agencies who have control over their own AI Providers can test GPT-5 on any of their OpenAI providers by clicking "Edit" and enabling the beta test.

You are now able to white label the domain of your client portal on the agency plan. Clients who you've given seats to will be able to access their client portal through the white label domain that you set up.

Not only has the abrupt nature of follow-ups been fixed, but you can now also give extra instruction as to how you want the AI to sound or behave when sending a follow-up message.

This week's focus was on performance of bookings and edge case bookings (like back to back bookings):

We've slowly increased custom scenarios because we know the more you use, the less accurate they become. We want to slowly increase the minimum as we are able to test performance. Enjoy having access to 5 custom scenarios now! We will continue to increase this number.

There are a considerable number of changes that have been made and are in the final phase of testing. There are a lot of front end changes that are waiting to be pushed until this testing is complete as well as some highly requested features. Next week will be a lot of fun.

For now, enjoy these updates!

This week was the last week focused solely on bug fixes. You'll see there have been many changes:

You'll notice a change in the conversations UI that allows you to clearly see how many objectives the Agent is working on and which nodes with clear icons. Hover over icons to see more information about them. You will also clearly be able to see booked appointments from within the UI and the time taken to generate a response. You'll also see icons for each knowledge library item that was pulled in for reference.

More changes coming to this screen next week will include the ability to click on icons to hyperlink to appropriate logs for quicker troubleshooting and more details in the section (like time the message was sent and estimated profit based on re-billing settings). You will also see follow-ups displayed here with the ability to modify and delete follow-up times that are scheduled.

The chat widget is now added as a channel. This refers to the SMS Widget that HighLevel supports. When a message comes in on this channel, we will responds via SMS.

Some users wish to allow Custom Scenarios to jumpt to a booking that has already been booked previously. Remember, by default Custom Scenarios can be entered multiple times. This allows you to let a contact book multiple appointments to the same calendar. Common for repeat customers.

Follow-ups now sound a lot less awkward. Next week we will release an update that allows you to edit for follow-ups, but this was the first step. It has taken so long for us to release this feature because of our hard mandate on no new features until we reached stability.

The client portal was previously showing these items in the sidebar that should not have been displayed.

The knowledge library is now much more reliable and we are open to feedback! We display when items are loading and when they have finished loading.

We now give you the ability to book your onboarding call(s) from within the CloseBot interface. This includes rescheduling instances of you missing your onboarding call.

The team has put in hard work over the past couple of weeks on general platform stability, but also some user interface updates. Many of the changes we are working on now are things that you don't see, but feel. You'll feel more and more that the platform operates faster and more reliably both for you and clients.

The knowledge base page got a much needed re-vamp. You'll see better filtering available for your knowledge base documents as well as column sorting. You'll also notice a column that shows the status of uploads.

- This means this item is still in process of being loaded into the AI memory

- This means this website is in the process of being scraped

- Indexed files have been loaded to memory and are ready to be attached to source(s)

Table views can now be sorted by column type. This is helpful if you're searching for recently modified job flows, for example.

More cleanup has gone into client views, removing items that they should not see (like knowledge base items that they aren't attached to).

To help with our push for stability, we have added some brand new conversation logs to show exact HTTP requests and responses. While we are still working on the UI for the logs, we wanted to push out quickly to help solve some of our small persistent bugs. This will be big for support but also for you!