When you upload content to Revian, it becomes part of the AI's knowledge base. But rather than just storing files, Revian transforms your content into something the AI can actually reason with and retrieve at the right moment.
Here's how that works in practice.
Answering questions with context
When a rep asks a question, the AI searches your content library for relevant information before responding. This means answers are grounded in your actual sales materials, not generic information.
Example
What's our implementation timeline for enterprise customers?
The AI searches your uploaded documents, finds the relevant section from your implementation guide, and provides a specific answer: "Enterprise implementations typically take 8-12 weeks, with the first 2 weeks focused on data migration and the remaining time on training and customization."
The AI cites its sources, so reps can click through to the full document if they need more detail. This builds trust and makes it easy to verify information before sharing with prospects.
Surfacing content during calls
During live calls, Revian listens for topics and questions that your content library can help with. When relevant content is detected, it appears in the AI assistant panel without the rep needing to ask.
Real-time suggestions include
- Case studies matching the prospect's industry
- Technical specs when product questions come up
- Pricing guidelines during negotiation discussions
- ROI calculators when value is being discussed
These suggestions are non-intrusive. They appear in a sidebar, letting the rep glance over when helpful without disrupting the conversation flow.
Battle card suggestions
When a competitor is mentioned on a call or in a chat, Revian automatically surfaces the relevant battle card. No searching required.
The AI detects competitor names, product mentions, and even indirect references like "we're also looking at your main competitor." It then pulls up talking points, differentiators, and common objection responses from your competitive content.
How it works
Prospect mentions: "We've also been talking to Competitor X about their solution."
The assistant immediately surfaces your Competitor X battle card, highlighting key differentiators, recent wins against them, and responses to their common claims.
Semantic search with vector embeddings
Behind the scenes, Revian uses vector embeddings to understand the meaning of your content, not just the words. This means searches work even when the exact terminology doesn't match.
For example, if a rep asks about "pricing flexibility" but your document uses the term "custom pricing packages," the AI still finds the right content because it understands the concepts are related.
Technical note: When you upload content, Revian breaks it into chunks, generates embeddings for each chunk, and stores them in a vector database. Queries are converted to embeddings and matched against this database using cosine similarity, returning the most semantically relevant content.
Deal-aware recommendations
The AI doesn't just search content in isolation. It considers the context of the current deal when making recommendations.
Factors the AI considers:
- Deal stage - Early-stage deals get discovery content; late-stage deals get case studies and ROI materials
- Industry - Content from similar industries is prioritized
- Deal size - Enterprise deals surface enterprise-focused materials
- Past conversations - Topics discussed in previous calls influence recommendations
- Stakeholders involved - Technical content when engineers are on the call, business content for executives
This means two reps asking the same question might get different recommended content based on their specific deal situation.
Tip: The more content you upload and the better you organize it with tags, the smarter these recommendations become.
Was this article helpful?