Blog
File Mention & HubSpot
We're excited to share some powerful new capabilities that we've added to Invicta AI to help you work more efficiently and effectively: 📌 File Mention You can now use the "@" symbol in your prompts to directly reference and pull in information from specific files, memories or integrations. This allows you to easily incorporate relevant data and context into your AI interactions. 🤝 HubSpot CRM Integration We've introduced a seamless integration with HubSpot CRM. You can now import your deals into Invicta AI, either in their entirety or filtered by specific stage or pipeline. Any updates made in HubSpot will automatically sync, ensuring your deal data in Invicta AI stays current. As always, we welcome your feedback. Please don't hesitate to reach out with any questions, suggestions, or concerns. Best regards, The Invicta AI Team
Claude-3, GPT-4 and LLMs are now all Free!
Hey there! 👋 We've been working hard behind the scenes to bring you our most exciting update yet. 🌟 Claude 3, GPT-4, and other AI Models - All Free! Access to Claude 3, GPT-4, and Mitral Large is now free. To manage this, we've introduced Compute Points. Each free user gets 50 daily compute points, equating to roughly 50 GPT-4 requests. Fully Customisable Web Widget 🌈 Your brand is unique, and your tools should reflect that. With our updated Web Widget, you can now tailor the look and feel to match your business colours perfectly, making every interaction distinctly yours. We transitioned to a Compute Points based system across the website💡 Minor Updates: 🌟 Our Landing Page and Documentation Page have received a fresh makeover. 📤 Removed file upload size limits. 📊 Enhanced Analytics now display token usage. 📱 Improved Mobile Compatibility for an enhanced experience on all devices.
Zendesk and 30% Better Knowledge Base Retrieval
Our AI agents just got smarter and are now fully equipped to integrate seamlessly into your customer support stack. Check out the latest enhancements: 🤝 Zendesk Integration Seamlessly integrate your AI agents into Zendesk to handle customer support queries across Email, Messenger, WhatsApp, Instagram, etc. This allows efficient, personalised responses powered by your AI agents. 📈 Enhanced Retrieval - Better by 30% We've boosted our AI's ability to pull the right information from your knowledge base—now 30% more accurate. For non-English content, the improvement leaps to 45%! Expect more on-point answers and smarter interactions from your AI agents. 💻 Improved Chat Widget Our chat widget now has a sleeker design and will not unnecessarily display files to users. Plus, it's more stable for a smooth experience. Other Enhancements: 🔧 Mobile experience is now more responsive and bug-free. 🛠 We've made infrastructure improvements for enhanced speed and the ability to scale with your business needs.
Google Gemini, Mistral, Profiles and more
Hi there, Let's dive into the latest updates on Invicta AI: 1. 🌌 Google Gemini & Mistral We've expanded our AI selection with Google Gemini and Mistral's newest AI models to offer you more versatility in your interactions. 2. 👤 Profiles Explore the Invicta AI community with the new profile feature. Check out what others are creating, or manage your own AI interactions seamlessly. 3. ⏸️ Pause Automations You're now in control with the ability to pause email automations whenever necessary. Minor updates: 🕒 Time-Aware Agents: Ask about past conversations or upcoming tasks with ease; our agents now understand the dimension of time! ✉️ Smarter Email Automations: They now include the sender's email and subject line for better context. 🔧 Bug Fixes: We've ironed out issues across several features, including chat uploads, memory functions, and the chat widget, and made enhancements to Claude 2.
Long-term AI Memory, Google Drive and Gmail
Happy New Year! 🎄 We hope your year is off to a fantastic start. To add to the excitement, we’ve introduced features to Invicta AI that we can’t wait for you to try. 1 ) 🧠 Your AI Agents can now learn from your chats • Create memories with AI: You can now save individual messages and entire chat threads into AI Agents’ Memory • Improves over time: As you chat with your AI Agents, they will become more helpful, remembering details and preferences to serve you better each time. 2 ) 🔄 Google Drive Sync • Link your Google Drive files directly to your AI Agents' knowledge base. • Any edits you make in Google Drive will be reflected instantly in the knowledge base. It's that seamless! 3 ) 📧 Gmail Integration • Email Automations just got easier. Connect Gmail with Invicta AI to streamline your email automations. Minor updates: • Fixed a bug where the last characters of messages were incorrectly sent as separate messages. • Improved the stability of email automation for improved and more reliable performance.
Threads, Chat Logs, Preferred LLM, and more
Here's what's new at Invicta AI, tailored for precision and convenience (12.12). 1 ) 🧵 Threads • You can now have multiple, individual conversation threads with AI agents • Email automations now possess thread memory, retaining the context of each email in a thread to ensure continuity in communication 2 ) 📊 Chat Log Analytics • Detailed analysis of how users interact with your AI agents and chat widgets. • User data is anonymised for privacy. Available to Business subscribers. Contact us. Minor updates: • ✨ Your LLM preference is now auto-saved across sessions; no need to reselect each time. • 🏠 You can now find your default AI agent atop your sidebar. You can change your default AI agent by going to the info window and clicking "Set Default" • 📎 Increased size limits for in-chat file attachments.
Analytics, Spreadsheet Automation, Typewriter Effect/Text Streaming, and more
Here's what we've rolled out at Invicta AI over the past week. 1 ) Spreadsheet Automations 📊 Upload spreadsheets, and your AI will generate content for each row, informed by its knowledge base. Try our templates for Candidate Ranking, Cold Email, and Essay Feedback. 2 ) Analytics v1 📈 View insights on how your AI agents are performing and get insights on your usage patterns. 3 ) Text Streaming/Typewriter (Finally!) ✍️ It's here at last – the much-requested text streaming feature. AI Agent responses now unfold in real-time, letter by letter. Minor updates: 🔔 • A revamped Knowledge Base UI 🗂️ • You can now set a default LLM for your chat widgets (Business) 💬 • Updated API docs 🧑💻 • Upgraded servers for a smoother and faster experience ⚡️ Take these features for a spin and let us know what you think!
Claude 2, In-Chat Attachments, Link Synchronisation
We are excited to announce various improvements to your Invicta AI experience. Here's what we have introduced over the last week (23.10). 1 ) Claude 2 and GPT-4 Turbo 🚀 Brush the dust off your website! You now have the ability to embed your AI agents as interactive chat bubbles or windows directly on your site. 2 ) Attach files to your conversations 📎 You can now upload CSVs into your AI agent’s knowledge base, empowering you to work with spreadsheets and task your AI agents with generating insights based on this data. 3 ) Knowledge Base Synchronisation 🔄 You now have the power to fully clone open-source AI agents. This allows you to customise and augment existing agents, as well as upload files directly into them. Minor updates: 🔔 • UI Redesign: Our user interface has been thoroughly updated for a more streamlined experience. 🖌️ • Forms: Variables in AI system messages now appear as forms at the beginning of the chat, allowing for efficient information gathering, such as contact details. 📝 • Chat Optimisation and Bug Fixes: Enhancements have been made to the chat functionality to ensure a smoother conversation flow. ⚙️ Enjoy these fresh updates and keep an eye out for more improvements from Invicta AI!
AI Automations - Your First Digital Employees
We're excited to announce our first step towards the future of work - AI Automations. Your AI agents can now act autonomously to execute tasks and workflows without your intervention. It's like having your first digital employees who never sleep! Introducing our very first AI Automaton - Email Automation. Your AI agents can now automatically answer all your emails, using the right information from its knowledge base. Demo: S8JDT.gif Here are some ways you can use it: ∙ Answer FAQs automatically. ∙ Arrange meetings for you. ∙ Find & summarise key emails Follow these 3 easy steps to set it up: ∙ Connect your Outlook (Gmail is coming soon!) ∙ Use an existing template or create your own AI agents ∙ And you're done! Try it out today and start making complex workflows simpler. If there are other workflows you want automated, let us know!
Embeddings and Widgets, CSVs, Cloning
We are excited to announce various improvements to your Invicta AI experience. Here's what we have introduced over the last week (23.10). 1 ) Embeddings and Widgets 📌 Brush the dust off your website! You now have the ability to embed your AI agents as interactive chat bubbles or windows directly on your site. 2 ) Upload CSV 📂 You can now upload CSVs into your AI agent’s knowledge base, empowering you to work with spreadsheets and task your AI agents with generating insights based on this data. 3 ) Cloning 🧬 You now have the power to fully clone open-source AI agents. This allows you to customise and augment existing agents, as well as upload files directly into them. Minor updates: 🔔 • References design has been refreshed; every source now appears as an individual bubble, and Web Links now redirect you directly to the source page - handy for widgets! 🔗 • Suggested Messages are now smarter and more contextually relevant than ever before. 💡 And finally, we've simplified and streamlined for maximum user-friendly experience: 🙌 • We've created a unified system message, so there's no need to switch between Text/Doc modes. When you upload documents to your AI agent, the context is automatically added. 📄 • Custom sliders have been deprecated in favour of variables for better compatibility with API-utilised AI agents and future automations. ⚙️ • We've removed max tokens as an option from LLM settings to keep it dynamic depending on the LLM in use. This update also preps us for the upcoming Claude2 with a 100k context window. 🚀 Enjoy these fresh updates and keep an eye out for more improvements from Invicta AI!