We’ll do out best to keep you updated on all that is happening behind the scene. Looking for to taking you along the journey. By @sup_nim + @amir__bio
- Better the input the better the output. So we’ve release a feature to do just that. Write you prompt and AI will ask further questions to help make the prompt better for you.
- We’ve made it easier to start a document chat. Via the document page and also from the create document form.
- A requested feature - have the option to turn the enhanced search on or off for your agents.
- Now you can have normal AI chats within aiagent.app, no need to leave to do one off questions (shoutout you guys for the feature request)
- Big one, we updated out UI massively to give agents and documents conversations their own space and added a section for all you uploaded documents. This update enables us massively to build out a proper robust system.
- Chat with CSV
- Source links exposed on responses, and is now shown on PDF downloads.
- After testing we’ve enabled Document Chat for free users too :)
- Woohoo! Now you can use URL as a form of data-source.
- Edit and archive Document Conversation
- Blog refinements
- FAQ refinements
- We’ve made it easier to edit, reorder and delete your tasks.
- Derive deep insights from your documents, at light speeds making you stay in flow. Now you can talk with your PDF’s.
- Increased token limit by 300%
We got feedback the response was getting cut off, we’re just changed the algorithm so it can now render more (300% more), hopefully, less will be cut off from now on. We’ll continue to monitor this and improve it further.
Share you’re Agent Conversations, the don’t even need to log in to see it.
We did user interviews to really understanding:
- what the people liked,
- what they want to see improved and
- what they’d like to see next.
It was nothing shy of incredible to talk you, thank you ♥️
Jodie, Enda, Steven, Stephen, Kate, Valentin, Marie, Nick, Anand, Ben, Kwaku, Erik, Josh, Daniel, Carroll
Export agent conversation in PDF, Image and text.
- Option to choose which GPT modal you want to use (GPT-3.5 or GPT-4)
- Failsafe image for profile picture, for when you don’t have one
- Other bug fixes and performance improvements.
- Now you can rename and archive conversations to keep your workspace clean and clear.w
- Run agent run. Now you have the option to have your task completed continuous without have to be there.
- Homepage refresh, now you have present to get you started.
A peek into the future of the product ;)
- Architecture update: We no longer automatically generate and reprioritise tasks after each task execution. This means that any new tasks that you add/edit will not be removed and the number of tasks to complete won’t keep on growing forever. As a result of this change, task execution is now 300% faster.
- Now you can manually add new tasks or let AI generate tasks for you
- Conversations are grouped based on when they were created:
- We now show which agents are running:
- Product planning, discussing what we’re going to build next
- Tooltip for goal writing and edit feature
- Pricing page improvements
- Now you can edit the suggested tasks from the Agent. (highly requested, we listened)
Note: our current agent architecture recreates the tasks and re-prioritises them on every loop. The very first task to be done next is not modified before it's executed but any other edits may be changed if the AI thinks it's not that relevant to the objective.
We'll have further customisation on this coming very soon!
- Chat styling enhancement — we’ve made it easier to read your content
- New week, new look 💅 - fresh home page!
- Refined subscription offerings to cater for much wider use cases
- Improved button loading UI (spinner colour is more visible on light mode)
- Fixed responses being slightly cut off on mobile
- Better UX for new user attempting to create a new agent before being logged in (the name and objective are retained and will be rehydrated after login redirect)
- Introducing ✨Enhanced Search ✨ Our Agent will now perform a more enhanced live search when it thinks it will help it achieve it’s goal
- Performance and stability fixes to our underlying AI Engines
- Record and show task creation history - you can now view which tasks were added as part of other tasks
- Set up sentry so we know when things go bad
- Significant architecture changes to handle a greater scale
- New Pricing design
- Fix loading state on mobile
- Research day
- Scaled our infrastructure to handle traffic
- Accessibility refinement on text
- Profile picture + accountOption dropdown
- Implemented Loading state when
- Add max length to agent name field
- Simplified our input fields
- Simplified our buttons component
- Implemented How it works section
- Implemented Testimonials
- Implemented Change log
- Implemented FAQ’s
- Server monitoring.
For another time, busy building.