User Interface

• ⚡️ The dashboard now shows scenarios detected if you have that scenario enabled in your bot settings

• ⚡️ UI allowing you to enable or disable different scenario checks along with the ability to put in your own custom scenario for detection. This is implemented in a way that it will not affect bot response speed at all.

Performance

• ⚡️ Situation detection added!!!

User Interface

• ⚡️ Users who have our Public Marketplace App installed have access to our free action library with the new free Conversation Summary action!

• ⚡️ Better warnings in the test portal so you know why the test portal stops responding

• ⚡️ Conversational rescheduling and cancellations now live for all paid plans. When the bot detects the contact needs to reschedule, it gets confirmation prior to cancelling the existing appointment. It will then "backtrack" to your latest accomplished calendar objective, uncheck it to show it's again incomplete, and try to complete it once again to conversationally schedule a new appointment

• ⚡️ Descriptions of what each AI stage is doing is now included when you click on a message and inspect each AI tab

• ⚡️ Transparent pricing now live, showing you exactly how much each generated message is costing you, whether you use CloseBot credits or your own API key. NOTE: message ID can now be found when clicking any message

Performance

• ⚡️ approximately 1.5 second response speed improvement per message for all users, made possible by cleaning up backend send message code and request caching.
• ⚡️ HIPAA accounts experienced a speed improvement of about 2 seconds per message. We no longer build always aware conversations for HIPAA accounts since that's not allowed by Lead Connector.
• ⚡️ Removed code implemented to block the bot sending hyperlinks if it's replying on the email channel so it can send hyperlinks via email, but as plain text if responding on any other channel
• 🪲 [ Bug Fix ] improved dynamic prompting around conversational booking to reduce chances of false confirmation if no appointment slots are available until tomorrow or further out if same day booking is requested.
• 🪲 [ Bug Fix ] Was showing credits being used even when they weren't actually being used (this was not drawing from your remaining balance, just showing credits used inaccurately). This has been fixed to show accurate credit usage.
• 🪲 [ Bug Fix ] different fallback behavior. Now if you try to route to a bot that isn't connected to the current Lead Connector account, there will be no bot response and a warning instead of falling back to your lowest bot number bot.
• 🪲 [ Bug Fix ] messages would sometimes send as duplicates when Lead Connector was under heavy load. Now with improved backend protection, this will not happen
• 🪲 [ Bug Fix ] different fallback behavior. Now if you try to route to a bot that isn't connected to the current Lead Connector account, there will be no bot response and a warning instead of falling back to your lowest bot number bot.
• 🪲 [ Bug Fix ] Duplicate bookings would happen when Lead Connector was under heavy load. Workaround for this on our end has been implemented
• 🪲 [ Bug Fix ] With recent changes to implement other features, aggression detection was overly sensitive. We optimized the prompt for this to bring the sensitivity back within expected range.

User Interface

• Public LeadConnector App published, allowing users to connect a sub-account with the CloseBot workflow already installed to an existing bot in less than 30 seconds

• Improved objectives page to allow for easier objective editing and view
• New User Interface to hard disconnect (uninstall app) from account if there was a connection issue and needs reconnected (now very rare)
• Able to Local test with live contacts to locally test how REAL booking works and field updating from the comfort of the CloseBot interface

• 🪲 [ Bug Fix ] Logs not showing for the Parse AI stage fixed
• 🪲 [ Bug Fix ] Search bar in dashboard message search was broken with a recent update and has been fixed
• 🪲 [ Bug Fix ] Advanced objective training wasn't showing up with new modal update. This has been fixed

Performance

• Email replies include original subject line now
• 🪲 [ Bug Fix ] ai off tag was not being auto added for new user accounts on manual reply with that setting turned on. This has been fixed.
• 🪲 [ Bug Fix ] Extra Prompt previously not used in Parse stage is now used once again, improving accuracy in field updating
• 🪲 [ Bug Fix ] Follow ups previously wouldn't work after longer than a 30 day wait, this has been resolved
• 🪲 [ Bug Fix ] Ability to select Company Name from auto update field section and the AI can update company name without issue. Previously you could select this, but bot wouldn't successfully update that field
• 🪲 [ Bug Fix ] Log saving errors have been resolved

User Interface

• Multi-account connecting to a single bot available for all non-subaccount CloseBot accounts

• Bots connected to multiple accounts will have an improved dashboard to allow them to view metrics per location

Performance

• Improved response speed and lower GoHighLevel disconnect rate from improving GHL token handling logic
• Improved backend caching to improve response speed
• 🗓️ Improved bot booking accuracy for those using restricted booking window (can't book x days out)
• Improved speed when saving new objectives on bots
• Latest Bot Message field updating with latest bot response as part of our provided fields
• 🪲 Previously sending your own message in the webhook was not working to override the inbound message contents as desired for GHL users. This has been fixed.
• 🪲 Bot objectives complete tag was not getting added sometimes if last objective was phone or email, this has been fixed

User Interface

• White Label connecting to sub-account routed back to CloseBot domain instead of white label domain. This is fixed

AI Performance

• Improvements to Anthropic follow up language and response quality
• Chat Widget auto route replies to default SMS channel
• Back to back booking bug fixes
• Latest response field added per request as contact custom field
• Summary field on contacts updated with latest executed summary
• AI Summary for long term memory improved to be more concise and include field update focus
• Email reply in thread bug fixes
• Email reply in thread during followups
• Log usage elapsed time in Usage logs upon clicking message
• Faster library references (needs optimized will improve over time)

User Interface

• 💎 Notification system to show bots that are falling back to use your CloseBot credits

AI Performance

• 💎 Email responses will now automatically be replied to in the same email thread and without using the Send Message workflow
• 💎 Bot responses to emails will automatically NEVER breakup messages or add typos to be more email-like in its reply.
• 💎 Now only workflows needed to make CloseBot function are... Process Message... that's it. We will clean up our snapshot next week, but no existing automations or snapshots will be broken because of this update (backwards compatible update)
• Fixed Anthropic followups sometimes responding in Spanish 🤷‍♂️
• Fixed Always Aware not pulling in all messages if Live Chat was closed by client and reopened for a new chat
• Fixed issue where rarely bot would respond with something like "I was not able to replace the requested..." when using our Word Swap feature.
• Fixed library referencing running twice if no applicable library items were found. This was sometimes slowing processing times for those using Library Items.

User Interface

• Able to select model priority from a drag and drop list in Bot Settings > General
• Test Bot page has changed to allow you to test your bot with OpenAI vs. Anthropic
• Toggle to turn off the default ON feature of always aware AI (discussed below)
• Changing between sub-accounts would sometimes show the incorrect bot in the dropdown compared to bot that was displayed in bot settings. This has been resolved.
• Removed ability to toggle allow fallback from User Settings page. This is now handled by modifying the model priority order within Bot Settings > General

AI Performance

• 🔥 [ New Feature ] Ability to choose bot LLM provider priority. The options include Anthropic (credits), OpenAI (credits), Anthropic (your key), OpenAI (your key). You can choose the order of these and even make it so credits aren't an option so your credits never get used. In this case, failed keys result in no AI response.
• 🔥 [ New Feature ] Always Aware AI which means the bot will now be aware of all recent messages to and from the contact (not just messages involving CloseBot). This means you can turn the bot off using ai off tag, have a conversation, turn the bot back on and the bot will be aware of what was said and update fields. This also means there's no longer a need for Override Messages
• More reliable and quicker auto-off on manually sent replies. Previously typo messages and broken up messages from the bot were falsely triggering the auto-off function.
• Updated Anthropic usage to use Sonnet 3.5 from previously used Opus

User Interface

• Bot settings page > Integrations would sometimes falsely show a bot as not being connected. This has been fixed

AI Performance

• [ New Feature ] Ability to allow a bot to stop automatically if a message is manually sent from a connected GoHighLevel account (paid feature)
• Stop Check will not trigger if the contact sends a number only as a response. This used to falsely trigger aggression from the bot if the lead responded with only a number
• Goals Accomplished field will be updated more reliably with updates to your Objectives in CloseBot and we will auto-remove "placeholder" objective that comes with snapshot
• Better API backoff for GoHighLevel requests during peak times, saving 1-2 seconds on response times during account-specific peak times

User Interface

• New News page where you can see news updates... like this one!
• Quicker site by improving badge load times (loaded periodically instead of every time)
• Banners to show when there's a LIVE event that you can join NOW

AI Performance

• Character limit increased if using the character_limit parameter of the web-hook
• Incremental backoff to combat GHL rate limits, resulting in more consistently quick response times during high volume messaging
• Booking error prompting changed to be less "forced" when there's a booking error (less awkward)
• Stop check was not being triggered, but is now fixed so it's being triggered again if you have that setting turned on
• Now able to select Website from the Field Update options to link to objectives
• Able to add to the follow-up prompting using the extra_prompt parameter available in the web-hook in the 1 Day Follow Up workflow
• If you change your GoHighLevel custom field(s) that may be linked to an objective, it will update automatically on our end as well as conversations happen instead of having to disconnect and reconnect that objective/field update to pull in changes.

We are headed to Dallas, TX in October as a Gold Exhibitor 👏

Check out other exhibitors here (notice we are the only AI-based bot product there as a Gold Exhibitor). Looking forward to meeting all of you there personally!

In October we will be close to the release of our re-vamped CloseBot website with much greater functionality... will share more about that with you later 😉