Changelog

Traditional support tools focus on deflection—putting walls between customers and help. They measure success by how many tickets they prevent, not by how well they actually solve customer problems.

But the CX leaders we talk to aren't impressed by deflection metrics anymore. As one recently told us: "Using AI in the wrong way is definitely risking customer happiness. I've been in demos that promised 90% ticket automation, and that's a quick red flag."

Ambient AI agents take a fundamentally different approach:

1. Prevention, not deflection - They identify and fix root causes of confusion
2. Contextual awareness - They understand what users are trying to do
3. Continuous learning - They get smarter with every interaction
4. Proactive guidance - They help before users get frustrated

Let's examine how each agent contributes to this connected experience.

Help Doc Agent: Documentation That Updates Itself

The documentation challenge has always been maintaining accuracy as products evolve. As one customer put it: "We just updated our docs last week and they're already outdated."

The Help Doc Agent transforms this dynamic by…

• Watching how users actually use your product
• Automatically evolving documentation to match reality
• Identifying gaps in content based on user searches
• Connecting related information across knowledge repositories

This saves your team time, of course, but it also improves accuracy and relevance. 

The Help Doc Agent improves key metrics like Customer Effort Score, Knowledge Freshness Rate, and Self-Service Success Rate.

Onboarding Agent: Personalized Guidance Without the Work

Traditional onboarding treats every user the same. Step 1, Step 2, Step 3—regardless of who they are or what they're trying to accomplish.

The Onboarding Agent works differently…

• It observes each user's behavior, like a video game learning your play style
• It offers guidance when someone seems stuck
• It creates different paths for different types of users
• It introduces advanced features only after basics are mastered

The results speak for themselves with metrics like reduced time-to-value, increased feature adoption, 30% reduction in training time, higher onboarding completion rates, reduced early-stage churn, improved NPS during onboarding, and fewer support tickets.

Support Agent: Helping Before Users Need to Ask

Most self-service options frustrate users more than they help. They force people through rigid flows that ignore their context and goals.

The Support Agent takes a different approach…

• It understands what users are trying to accomplish right now
• It offers relevant help when someone seems confused
• It points people to the right resources based on their specific situation
• It solves common problems without making users wait for a human

Companies like Smokeball have seen remarkable improvements, with a 92% self-service rate and support NPS jumping from 60 to 77.

Technical Support Agent: Expert Help Without the Wait

Your most technical problems often wait the longest for help because they need your limited specialists who are already overwhelmed.

The Technical Support Agent changes this by…

• Solving technical issues that normally need escalation
• Walking users through complex setup and configuration
• Offering troubleshooting help that understands their specific situation
• Spotting patterns in difficult support cases

This transforms metrics like L1 & L2 ticket volume, escalation rate, technical resolution time, L2 backlog size, technical specialist productivity, and implementation time.

Role-Based Agent: Tailored Support for Every User Type

Not all users are the same, so why give them all the same help?

The Role-Based Agent creates personalized experiences by:

• Building separate knowledge bases for different user groups automatically
• Offering personalized help for different types of users
• Using the right terminology and examples for each industry
• Showing features relevant to specific pricing plans

This improves CSAT/NPS within targeted segments, account retention, upgrade conversion rates, time-to-value by segment, and cross-sell/upsell success.

How These Agents Work Together

What makes these agents truly powerful is how they work together as a system…

1. Shared Context Understanding - When a user interacts with your product, all agents share the same contextual understanding of who they are, what they're trying to do, and where they're struggling.
2. Collective Learning - Insights discovered by one agent become immediately available to all others. If the Support Agent identifies a common confusion point, the Help Doc Agent can automatically create relevant documentation.
3. Seamless Transitions - As users move through different phases of their journey, the experience remains consistent. The Onboarding Agent smoothly hands off to the Support Agent when needed, with no context lost.

Unified Analytics - Instead of siloed metrics, you get a holistic view of the entire customer experience, revealing patterns and opportunities that would otherwise remain hidden.

This February, we've released several major updates that make Brainfish even better at understanding and helping your users. These improvements focus on making your product naturally easier to use through ambient AI and more seamless assistance.

Self-Serve Agents: Context-Aware Support Right in Your Product

We all know what happens when users get stuck – they either wait for support or try to figure things out themselves, often getting more frustrated along the way. That's why we built Self-Serve Agents that actually understand your product and give users the right answer right away.

When someone clicks the Brainfish widget, they get immediate help that's specific to where they are and what they're trying to do. No more digging through help docs or waiting for email responses. Just quick, reliable answers that help users keep moving forward.

What makes this different?

• Your users get help instantly, any time of day. No more Monday morning ticket backlogs or frustrated users waiting for answers.
• Your support team can focus on the complex stuff. Instead of answering the same questions over and over, they can tackle the interesting challenges that really need their expertise.
• Every answer is tailored to the user's situation. The agent understands where they are in your product and what they've tried, so answers are always relevant.
• Getting started takes minutes, not months. No complicated setup or maintenance needed.

Learn more here.

API Connections: More Data Sources to Answer User Questions

Brainfish now connects with more of your product's data and systems, helping users get better answers and complete tasks without leaving their workflow.

• Connect Your Systems: Pull in data from more sources to give users complete, accurate answers
• Take Action: Help users actually solve problems, not just read about solutions
• Embed Anywhere: Add help exactly where users need it in your product

Follow-Up Improvements: Ensuring User Success

The new Follow-Up Widget helps ensure users don't just get answers – they actually accomplish what they set out to do.

• Check Progress: See if users successfully completed their task after getting help
• Guide Next Steps: Show users exactly what to do next with clear action buttons
• Stay in Context: Suggest relevant actions based on what the user is trying to accomplish

What's Next

These updates represent a major step forward in our mission to make products naturally easier to use. We're already working on more improvements to help your users succeed with even less effort.

Have questions about any of these updates? Our team is here to help you make the most of these new capabilities.

We've added powerful new ways to organize your help content and create richer documentation. These updates make it easier for users to find exactly what they need while giving you more flexibility in how you present information.

Help Content Organization

Knowledge Grouping lets you categorize your help content into logical groups, making information more discoverable and relevant for users.

• Create custom groupings to match how users actually think about your product
• Improve search accuracy by providing better context around related content
• Organize collections and sources without disrupting existing structures
• Give users faster access to the specific information they need

Learn more here.

Enhanced Article Embedding

New iFrame embedding lets you integrate external content directly into your help articles, creating more comprehensive documentation.

• Add interactive content with a simple /iframe command
• Preview embeds before publishing to ensure perfect display
• Maintain full control over how external content appears
• Create richer, more engaging help documentation

Learn more here.

These updates continue our focus on making help content more accessible and useful. By combining better organization with richer content capabilities, we're making it easier for users to find and understand the information they need.

Instant Website Integration

Connect your entire website to Brainfish with a single click. No more manual content migration – just point and deploy. Keep everything fresh with on-demand content syncing.

• Immediate deployment instead of weeks of content migration
• Preserve your existing content structure and organization
• Keep your knowledge base current with manual sync options
• Start getting user behavior insights from day one

Advanced Search Intelligence

We've added GPT reranking technology that carefully evaluates search results before generating answers. Super admins can customize this feature for each customer's specific needs.

• Higher accuracy for complex product questions
• Customizable search precision for different content types
• More relevant answers, especially for technical queries
• Note: May add a brief processing delay for enhanced accuracy

Context-Aware Responses

Enhanced context validation ensures users get answers specific to their exact situation, not just similar matches.

• Prevents confusion between related but distinct features
• Particularly valuable for products with nuanced functionality
• Reduces support escalations from ambiguous answers
• Perfect for companies with complex product ecosystems

Behind these seemingly simple updates lies a lot of engineering that makes self-service feel natural. We're continuing our mission to help users get the right answers instantly, while giving support teams deeper insights into user needs.