Using RAIN Intelligence on 8020CRM Records
- Apr 30
- 6 min read
Overview
What Is RAIN Intelligence on Records?
As of the April 2026 release (v1.44 / RAIN OS v0.10), RAIN Intelligence is available directly on every Lead, Opportunity, and Property Transaction record page in 8020CRM. You do not need to navigate to a separate tool or tab. It lives on the record you are already working.
This surface is called the Record Copilot. It gives you a context-aware AI assistant that already knows everything on the record - seller information, activity history, stage and status, field values, notes, and call transcripts. Ask it questions in plain language and it answers based on what is actually in your Salesforce org.
Before You Begin
What You Need
RAIN Intelligence must be active in your org. Contact your System Admin if you do not see it on record pages.
Your org must have platform-provided RAIN credits or a connected API key configured in the GetReal Account.
Standard user access to the record you are querying. RAIN follows Salesforce permissions and will not surface data you do not have access to.
Records with more complete data - filled Pillar fields, logged calls, detailed notes - will return more accurate and useful responses.
Why It Matters
Before Record Copilot, getting a quick read on a deal meant scanning the Timeline manually, checking each field, and pulling up notes one by one. For an experienced team member that takes two to three minutes per record. Multiply that across a morning of follow-up calls and the overhead adds up.
Record Copilot compresses that research into seconds. More importantly, it helps your team ask better questions - not just what happened on this lead, but what should happen next, what risks exist, and what the right message is for this specific seller at this specific moment.
It also creates a consistent baseline across your team. Instead of every person making a different judgment call on follow-up timing and tone, RAIN provides a shared starting point informed by every piece of data on the record.
Two RAIN Intelligence Surfaces
RAIN Intelligence appears in two places in 8020CRM. Understanding the difference helps you use the right one for the right task.
Surface | What It Is and Who Uses It |
Record Copilot (on the record page) | Context-aware AI that already knows everything on the specific record you are viewing. Six built-in prompts focused on record-level tasks: summarize, score, find risks, draft follow-up, next action, visualize. Available to all roles for daily use. |
RAIN Intelligence Workspace (standalone app) | Broader AI workspace accessed from the utility bar or App Launcher. Twelve prompts covering pipeline analysis, campaign design, automation building, UI optimization, and admin-level org management. Different prompts are relevant to different roles. See the separate article: Using the RAIN Intelligence Workspace. |
This article covers the Record Copilot. For the full RAIN Intelligence Workspace, see Using the RAIN Intelligence Workspace in 8020CRM.
The Six Record Copilot Prompts
When you open RAIN Intelligence on a record you will see six pre-built prompt buttons. Each is designed for a specific task your team does daily. All six are context-aware - they run against the specific record you are viewing, not your pipeline in general.
Prompt | What It Does |
Summarize This Record | Returns a plain-language summary of the record - seller information, property details, current stage, recent activity, and logged notes or call transcripts. Use this at the start of a call to get up to speed in under 30 seconds. |
Score This Record | Evaluates the lead or deal based on available data - motivation signals, qualification completeness, timeline, and activity pattern - and returns a score with a brief explanation. Use this to prioritize your call list. |
Find Risks | Scans the record for flags that could indicate a deal at risk - stalled stage, missing qualification data, seller hesitation signals in notes, long gaps between activities. Use this before committing resources to a deal. |
Draft Follow-Up | Generates a follow-up message - email or SMS - based on the last recorded interaction. The draft is personalized to the seller using record details and conversation history. Use this to save drafting time and keep outreach consistent. |
What Should I Do Next? | Recommends the single most important next action based on where the record is in the process, when the last activity occurred, and what the seller has communicated. Use this when you are unsure how to prioritize. |
Visualize This Record | Returns a chart or structured visual summary of deal activity, timeline data, or stage progression. Use this during team reviews or when presenting deal status to leadership. |
These are starting points, not the only things you can ask. After running a prompt, continue the conversation - ask follow-up questions, request clarifications, or ask it to adjust a draft it generated.
CONVERSATIONAL USE: You can type any question directly into the RAIN Intelligence input on the record. Examples: 'Does this seller have an equity situation worth pursuing?', 'How long has this lead been in the Working stage?', 'What did the seller say about their timeline on the last call?' RAIN will answer from the record data.
Approving and Rejecting Next Best Actions
In addition to answering questions, RAIN Intelligence surfaces Next Best Action suggestions directly on the record as banners or inline prompts. These are separate from the six built-in prompts, but part of the same system.
When a Next Best Action appears, you have three options:
Approve: The suggested action pre-fills the Activity Composer. If the suggestion is to send an SMS, the Composer opens with the message body written. If it is an email, the subject and body are pre-populated. If it is a task, the task form is pre-filled. Review, adjust if needed, and execute.
Reject: Dismisses the suggestion for this cycle. RAIN will not suggest the same action again until the record state changes.
Dismiss: Clears the suggestion without a strong signal. RAIN may resurface a similar suggestion at the next interval.
ADMIN NOTE: Whether Next Best Actions fire automatically or require manual approval depends on the confidence thresholds and execution delays your System Admin has configured. As of April 2026, these are configurable per action type - SMS, email, task, event, and voice call each have their own settings. Ask your System Admin how your org is configured.
Voice Conversation Mode
If your org has voice conversation mode enabled, you can talk to RAIN Intelligence instead of typing. Look for the microphone icon in the RAIN Intelligence panel on the record. Speak your question and RAIN will respond.
Voice mode is useful during a live call when your hands are occupied or when you want a quick answer without breaking your conversation rhythm. Same RAIN Intelligence, same data, same responses - voice-triggered instead of typed.
CURRENT LIMITATION: Voice mode in April 2026 is not fully duplex. You speak, RAIN processes, then responds. It does not support back-and-forth interruption the way a natural phone conversation would. Full duplex support is on the product roadmap.
How RAIN Follows Your Permissions
RAIN Intelligence operates within the same Salesforce security model as everything else in 8020CRM:
If you cannot see a field in Salesforce, RAIN cannot tell you its value.
If you do not have access to a related record, RAIN will not pull data from it.
If financial fields, commission data, or ROI calculations are restricted for your profile, RAIN will not surface them in its responses.
This is by design. Your administrators control what each role can see, and RAIN respects those boundaries automatically.
Things to Know
A few behaviors worth understanding as you use RAIN Intelligence on records:
Custom boolean fields with action-oriented names - for example, a field named Send_SMS or Call_Now - may generate unexpected Next Best Action suggestions. RAIN reads boolean fields and can interpret them as action signals. If suggestions are appearing that do not make sense, flag them to your System Admin. They can filter out fields that should not generate suggestions.
RAIN Intelligence suggestions are assigned to the user who triggered them, not the record owner. If you run a prompt on a colleague's record, the resulting suggestions and any auto-applied field updates are attributed to you.
Source citations appear below RAIN Intelligence responses so you can see where RAIN found its information within your org data.
Attachments and screenshots you include in a RAIN conversation persist with the message - you can scroll back to find them later.
Best Practices
Log every call, note, and interaction in the Activity Composer. RAIN reads the Timeline to build its context. Sparse records produce sparse answers.
Keep status and stage updated in real time. RAIN uses these as primary signals for scoring, risk assessment, and next best action recommendations.
Run Summarize This Record before every outbound call. Ten seconds eliminates the chance of referencing stale information.
Use Find Risks before moving a deal forward. Easier to address a red flag before commitment than after.
Treat Draft Follow-Up output as a starting point. Read it, adjust for tone and context, then send. Never send AI-generated content without a human review.
If a Next Best Action does not look right, reject it - do not dismiss it. Rejections provide a stronger signal to the system that the suggestion was off. Your System Admin can use rejection patterns to refine agent configuration over time.
SOURCE OF TRUTH: RAIN Intelligence is only as accurate as the data on the record. A clean, well-logged record with current status, complete Pillar fields, and detailed activity notes produces significantly better responses than a sparse record. Your data discipline directly determines the quality of what RAIN returns.
Comments