Fini Product Features
Jul 1, 2024
Zuzanna Ostrowska
IN this article
This July, we’re bringing powerful new features to enhance AI-driven support and streamline customer interactions. Fini Knowledge Items ensure up-to-date support information, while Chat to Knowledge Base automatically converts conversations into knowledge. Flows (mini-agents) guide users through structured processes, and our Product Recommendation Engine personalizes shopping experiences effortlessly. We’ve also introduced Modified User Roles for better access control, expanded support with a HubSpot Integration, and improved Conversational Flow for smoother interactions. Looking ahead, User Attributes will enable even more personalized AI responses—stay tuned for what’s next!
Welcome back to our Monthly Product Update! This July we’re very excited to share that we’ve launched:
📖 Fini knowledge Items
💬📚Chat to Knowledge base
🔀Flows (mini-agents)
🔝Product recommendation engine
🧑💻Modified User roles
🤝Hubspot integration
⚒️Improved conversational flow
📖 Fini Knowledge Items
Keeping your customer support information up-to-date is a big challenge. Often, information gets outdated, leading to slow and inefficient support. This can frustrate your customers and waste your team’s time. Fini Knowledge Items are designed to solve this problem by making it easy to create, maintain and update your customer knowledge using AI .
Our vision is to create a self-updating knowledge base using your previous customer chats, ongoing chats etc., and knowledge items will act as the building block for our future vision
If you’re interested in discovering the feature, read more here or log in to your Fini account and check it out!
💬📚 Chat to Knowledge base
Chat to Knowledge Base leverages your existing customer conversations from platforms like Zendesk, Intercom, or Front to automatically create and enrich your knowledge base. This new feature not only reduces the manual effort required to input data but also accelerates the setup and training of our models, ensuring a seamless integration experience for you.
To gain some better understanding of the feature check out our blog.
🔀 Flows
This feature allows structuring conversations into predefined, logical sequences that guide users through specific processes or decision trees through micro-prompting and chaining multiple agents together. This ensures a smooth, intuitive user experience for most sensitive inquiries (refunds, cancellations, you name it). Think of this as mini agents specialized in specific tasks eg: you can create a highly specialized cancellation expert that will help you with retention, a payments expert that helps with chargebacks etc.
Curious to learn more about flows? Check out our blog post here
🔝 Product recommendation engine
Imagine someone wants a phone within a certain price range, a stylish red dress under $50, or a hotel in a specific city for their next vacation. They simply type in their requirements, and Fini quickly gives them the best options. It’s as easy as having a conversation with a knowledgeable friend who knows your entire product catalog.
Read more about this feature here.
🧑💻 Modified User roles
You’re now able to differentiate the access level in Fini platform between:
Agent: access only to Ask Fini
Member: access to: Ask Fini, Chat History, Analytics, Agent QA tool (Paramount)
Admin: access to all tabs within Fini
🤝 Hubspot integration
Our headless AI Support System just got another possibility of deployment! Now you can use Fini on Hubspot! (In addition to: Zendesk, Intercom, Front, LiveChat)
⚒️ Improved conversational flow
You asked and we listened! Fini is now able to maintain a conversation (within chat or whatsapp) thanks to the changes to the context window. For a more detailed understanding check out this video.
⏭️🔮 Coming soon...
🔍 User Attributes
This upgrade will help Fini change its replies based on user details. By customizing interactions to fit each person's needs and situation, we aim to provide a more personal and effective user experience.
And we’re not FINIshed… 😉 Stay tuned for more amazing features coming your way!
Knowledge Items
1. What are Fini Knowledge Items and why are they important?
Fini Knowledge Items are structured, modular pieces of information that serve as the building blocks of your support content. Instead of relying on static help articles or dumping raw chat logs into AI, Knowledge Items ensure your AI agent always has access to accurate, up-to-date responses. They’re especially useful for reducing manual upkeep and enabling AI to scale intelligently.
2. How do Knowledge Items improve support accuracy over time?
As conversations evolve, outdated or incorrect knowledge can creep in. Fini Knowledge Items are built to be dynamically updated from ongoing interactions, manual edits, or synced chat data, ensuring that the AI only relies on the most relevant and verified support content.
3. Can teams manually edit or prioritize Knowledge Items?
Yes, support teams can add, update, remove, or reorder Knowledge Items at any time through the Fini dashboard. This gives human agents control over what the AI references and allows for faster adaptation to product changes or user feedback.
4. How do Knowledge Items connect to other Fini features like analytics or Flows?
Knowledge Items directly power Flows by providing reliable information for each step in the journey. They also feed into Analytics to help measure content effectiveness and coverage per category, identifying gaps and optimization areas.
5. Can Knowledge Items support multiple languages or tones?
Yes, each Knowledge Item can be customized with specific instructions, including tone preferences, formatting, or translation logic. This makes it easy to create multilingual, brand-aligned experiences with minimal duplication.
Chat to Knowledge Base
6. What is Fini's "Chat to Knowledge Base" feature?
This feature turns historical conversations from platforms like Zendesk, Intercom, or Front into Knowledge Items. It allows teams to auto-generate support content based on what agents have already been saying successfully.
7. How does Chat to Knowledge Base ensure quality control?
Fini filters out incorrect, repetitive, or unclear responses during the conversion process. The AI selects only useful, high-performing segments from past chats that clearly solve user problems.
8. Can teams approve or edit auto-generated knowledge from conversations?
Yes, the content created through Chat to Knowledge Base can be reviewed and edited before publishing. This ensures human oversight and brand consistency while still accelerating content creation.
9. Is it possible to generate Knowledge Items in real-time from ongoing chats?
Absolutely. Teams can trigger generation after a conversation ends or on-demand to continuously improve support coverage without waiting for quarterly content audits.
10. How fast can teams build an initial knowledge base using past tickets?
In most cases, teams can build a robust base of 50–100 Knowledge Items within days by syncing past ticket data and approving high-value answers. It dramatically reduces setup time for new AI agents.
Flows (Mini Agents)
11. What are Flows in Fini?
Flows are pre-configured mini-agents that guide users through structured, multi-step journeys. They're perfect for complex processes like cancellations, returns, onboarding, or upselling, and allow AI to stay context-aware throughout.
12. How are Flows different from regular AI conversations?
Regular AI responses are reactive and driven by single prompts, while Flows use branching logic, multiple prompts, and memory to guide users through decision trees tailored to their needs.
13. Can Flows be customized for different customer segments?
Yes, Flows can include logic based on user attributes like account type, subscription tier, churn risk, or purchase history. This enables precise targeting and personalization at scale.
14. What kind of business outcomes can Flows improve?
Flows increase retention (by offering personalized cancellation alternatives), reduce repetitive tickets (e.g., for order tracking), and increase revenue (by surfacing relevant upgrades or add-ons).
15. Can Flows escalate to a human agent mid-conversation?
Yes, if the Flow hits a blocker or detects sentiment like frustration, it can trigger an escalation protocol or leave a note for a human agent to take over seamlessly.
Product Recommendation Engine
16. What does the Product Recommendation Engine do?
Fini’s Product Recommendation Engine allows AI agents to suggest relevant products based on natural language queries. For example, a user might ask for “running shoes under $100,” and Fini will deliver targeted, accurate recommendations.
17. How does Fini access and understand the product catalog?
Fini connects to your product database or catalog feed and indexes the metadata (e.g., price, category, description, stock). It uses that data to interpret user preferences and apply search filters.
18. Can the recommendations include visual previews or purchase links?
Yes, product suggestions can include images, pricing, links to product pages, and even discount codes or urgency messages depending on configuration.
19. How does this feature help increase conversion rates?
By answering pre-purchase questions within the support channel, Fini shortens the decision journey and captures sales that might otherwise require human follow-up or lead to drop-offs.
20. Can recommendations be tailored based on previous customer purchases?
Yes, with user attribute integration, Fini can suggest complementary items based on what the customer has already bought or browsed.
HubSpot Integration
21. What does the HubSpot integration enable?
Fini integrates with HubSpot to sync customer data, push tags, categorize tickets, trigger workflows, and power AI responses based on CRM attributes. This turns Fini into an AI-first extension of your HubSpot support flow.
22. Can Fini operate across HubSpot email and chat channels?
Yes, Fini works across both, handling ticket replies, live chat responses, and internal note generation for agents inside the HubSpot Service Hub.
23. How does Fini trigger workflows inside HubSpot?
AI agents can tag or categorize conversations based on intent, resolution, or urgency—automatically triggering HubSpot workflows for follow-up, pipeline updates, or support escalations.
24. What’s the typical setup time for HubSpot integration?
Most customers get fully live in under a day, thanks to Fini’s no-code setup and native support for HubSpot APIs and chat widgets.
25. Can Fini auto-update contact records in HubSpot based on interactions?
Yes, Fini can enrich user profiles with tags, sentiment, conversation outcomes, and other metadata to make contact records more actionable.
User Roles and Access Control
26. What are the different user roles in Fini?
There are three main roles: Agent (access to Ask Fini only), Member (access to Ask Fini, chat history, analytics, QA tool), and Admin (access to all features including configuration and knowledge management).
27. Can permissions be customized for each team?
Yes, Fini supports granular access control, allowing you to assign roles based on department, responsibilities, or region.
28. How do roles affect visibility into customer conversations?
Only those with appropriate roles (Member or Admin) can view past conversations, analytics, or configuration settings. Agents only see the Ask Fini interface.
29. Is role setup integrated with SSO or identity providers?
Yes, enterprise customers can configure roles through SSO or identity management systems like Okta or Google Workspace for secure onboarding and offboarding.
30. What happens if a user is downgraded from Admin to Agent?
Their access is immediately restricted to the appropriate tabs, and they lose the ability to edit knowledge, Flows, analytics, or configurations.
General Product Enhancements
31. How does the new HubSpot integration differ from Intercom or Zendesk?
While core features like auto-responses and knowledge integration remain consistent, HubSpot’s deployment focuses more on CRM automation, enabling Fini to act on contact attributes, lifecycle stages, and marketing tags.
32. What improvements were made to conversational flow this month?
Fini now maintains more persistent context across conversations, meaning it remembers key details even in longer interactions, including channel-specific adjustments for tone, sequence, and sign-off logic.
33. Can I run A/B tests on Flows or Knowledge Item effectiveness?
Yes, Fini Analytics lets you compare performance across categories, messages, and Flows to see which are driving the highest resolution rates or user satisfaction.
34. How frequently should Knowledge Items be reviewed or updated?
It’s best practice to review high-volume Knowledge Items biweekly and low-volume ones monthly, especially after product updates, launches, or policy changes.
35. How do I know if a Flow is underperforming?
Analytics will show drop-off rates, escalation frequency, and resolution scores for each Flow. You can use this to fine-tune prompts or add steps.
Looking Ahead
36. What are “User Attributes” and how will they improve personalization?
User Attributes include customer-specific details like plan type, usage frequency, LTV, churn risk, or last purchase. These allow Fini to personalize every reply and route users through the most relevant Flow or resolution path.
37. What is “Chat Injection” and why does it matter?
Chat Injection lets you customize tone, formatting, or logic based on the support channel. For example, responses on WhatsApp may be casual and emoji-friendly, while email replies are more formal with sign-offs.
38. What is “Conversational Flows v3” and what can we expect?
The next version of Flows will allow AI agents to take real actions—like pausing subscriptions, updating shipping info, or escalating internally—based on user context and knowledge.
39. Can I preview upcoming features like the search bar?
Yes, early beta users can already test the new search bar widget, which enables users to find answers before submitting a ticket—reducing volume and improving CSAT.
40. How do I request early access to new features like Flows v3 or Fini Search?
You can contact the Fini team at founders@usefini.com or book a demo with us to join beta programs or pilot new deployments.
AI Customer Success Manager
