Every Lead and Opportunity scored 0–100 — with a written reason and a recommended next step, based on your own won/lost history. No Einstein add-on, your own AI provider key.
The AI score card on a record — score, plain-language analysis, recommendation, and one-click Next Steps and Draft Email actions.
No black box. Every score is produced from context you control and comes back with the reasoning written out — so reps can read it, trust it, or challenge it.
AI Engine reads only the fields your admin approved for the Lead or Opportunity — standard and custom. PII fields like email and phone stay out unless explicitly allowed.
A sample of your recently won and lost records is added for pattern matching — so the score reflects what actually closes in your pipeline, not a generic industry model.
Your AI provider — OpenAI, Claude or Azure OpenAI, via your own key — weighs fit, data completeness, engagement signals and deal momentum against those patterns.
The score (0–100), label and written reason are saved to the record, and the card shows a recommended next action. Scoring runs automatically, on demand with Generate AI Score, or in scheduled batches.
{
"score": 82,
"label": "Hot",
"confidence": "High",
"reason": "Strong fit with your recent wins in manufacturing: budget confirmed, decision maker engaged, active in the last 7 days. Missing: agreed next meeting.",
"recommendation": "Propose a pricing call this week while engagement is high."
}The actual JSON returned for every scored record — written to AI Score, AI Score Label and AI Score Reason fields on the Lead or Opportunity.
Since Spring 2026, Einstein Opportunity Scoring is included for all Sales Cloud users — and for some teams that’s enough. Here’s where the two approaches genuinely differ.
| AI Engine Scoring | Einstein Lead & Opportunity Scoring | |
|---|---|---|
| Opportunity scoring | 0–100 score with written reason and recommended next step | Score with factor cards; included for Sales Cloud users since Spring 2026 |
| Lead scoring | Included — same engine, same flat price | Requires Enterprise Edition or higher plus the Einstein AI add-on |
| Salesforce editions | Enterprise, Unlimited, Developer (on Professional, contact us first) | Higher editions; Professional not supported |
| How the score is explained | Plain-language reason + concrete recommendation a rep can act on | Top contributing factors shown as cards |
| Scoring model | Your own won/lost records as live context, plus your sales playbook via custom prompts | Predictive model trained on your historical data; global model until enough records accrue |
| AI provider & data control | Your choice — OpenAI, Claude or Azure OpenAI, via your own API key and Named Credential | Salesforce-managed models |
| Custom objects & fields | Admin-configured field allowlist, standard and custom fields | Standard Lead/Opportunity fields, limited customization |
| Beyond scoring | Same package includes assistant chat, summaries, email drafting, next steps, meeting prep | Scoring only; other AI features are separate products |
If you’re on Enterprise Edition and only need opportunity scores, try Einstein first — it’s included. AI Engine is for teams that want lead scoring without the add-on, written reasoning, or one package that also drafts, summarizes and answers questions.
AI Engine keeps the experience close to Salesforce users, admins and business teams.
Generate practical score labels and explanations to help teams focus on the right deals.
Identify high-potential opportunities, stale deals and missing data signals faster.
Surface red flags, weak qualification, missing stakeholders and next-step gaps.
Adjust scoring logic, prompt templates and available fields based on your Salesforce process.
Use approved Salesforce context, configurable prompt templates and your preferred AI provider to support everyday sales and service work.
Designed for Salesforce admins and business systems teams, with clear configuration instead of hard-coded one-off automations.
Choose the provider that fits your security, compliance and cost requirements.
Use standard and custom Salesforce objects, fields and child records depending on your use case.
AI Engine builds a context from the record fields your admin approved, adds a sample of your recently won and lost records for pattern matching, and asks your AI provider for a 0–100 score. The response includes a label, a written reason and a recommended next action. The score, label and reason are saved back to the Lead or Opportunity; the recommendation is shown on the AI Score card.
Not the Einstein add-on. Einstein Lead Scoring requires Enterprise Edition or higher plus the Einstein AI add-on; AI Engine runs scoring through your own AI provider key on Enterprise, Unlimited and Developer editions.
It can be — but it doesn’t have to be. Since Spring 2026, Einstein Opportunity Scoring is included for Sales Cloud users, and if that covers your needs, use it. AI Engine adds lead scoring without the add-on, written reasoning instead of factor cards, and your choice of AI provider.
Three ways: automatically when a new record is created or a key field changes (if the admin turns this on), on demand with the Generate AI Score button on the record, or in scheduled batches.
Yes. Admins configure exactly which standard and custom fields are included in the AI context. PII fields like email and phone are excluded unless explicitly allowed.
Only the admin-approved fields are sent to the AI provider you configured — via a Named Credential with your own API key. AI Engine does not store the full prompts and answers; automatic scoring writes only the score, label, reason and scoring date to the record.
Configure AI scoring for Leads and Opportunities — activation, won/lost history, your own rules, automatic scoring and scheduled batches.
Choose scoring context fieldsPick exactly which fields the AI reads when it scores a Lead or Opportunity, and decide whether to verify the company on the web.
The AI Score cardRead and refresh the 0-100 AI score on Leads and Opportunities, and see why the record got it.