AI Autofill: automatically fill CRM fields from your deal timeline or the web

Your CRM is only useful if the data in it is current.

Keeping fields up to date — next steps, deal status, company details, lead research — usually means time you'd rather spend on actual work.

Autofill does it for you, reading what's already in your deals and conversations or turning to the web to find what isn't.

‍

Keep your pipeline current without manual data entry

Every deal generates information that belongs in your CRM — agreed next steps, key dates, lead context, deal status — but it usually stays buried in the thread.

Autofill reads your email history, call notes, and meeting records and enters that information into the right fields, so your pipeline reflects what's actually happening without anyone having to update it manually.

• Next steps stay current after every email. Autofill reads the most recent thread and fills the field — no manual notes, no follow-up lag.
• Leads get classified on arrival. When a new inquiry comes in, Autofill reads it and populates your pipeline with the right value, respecting your defined classification options for clean, reportable data.
• Key dates make it out of email threads. Contract expiry, trial end dates, stated timelines — Autofill finds them in emails and writes them into structured date fields.
• Lead context is captured at the right moment. Pain points and use cases from the first inquiry go into a field immediately, ready to filter or report on later.

‍

Research leads without leaving your CRM

For data that doesn't exist in your inbox — company details, lead intelligence, enrichment — Autofill can search to the web to find it.

Give each field its own research instruction and Autofill runs it on one record or in bulk across your entire pipeline, writing results directly into your existing columns.

• New leads arrive enriched. Company size, industry, funding stage, and HQ location are already filled in by the time you open the record.
• Every lead gets scored against your ICP. Autofill evaluates each lead against your criteria and returns a score with a sentence of reasoning, so qualification is consistent across your team.
• Outreach openers write themselves. Autofill pulls a recent company milestone or news item and drafts a personalized first line — ready to drop into your mail merge.
• Buying signals are visible before the first call. Tech stack, recent hiring patterns, and other public signals are surfaced into structured fields before anyone picks up the phone.

Autofill does this inside the pipeline where your team already works, using the columns you've already built. This means you can eliminate standalone data enrichments tools to consolidate your tech stack.

‍

Get precise results with custom instructions

Custom instructions let you write a detailed prompt for how a field should be filled — not just where the data comes from.

Custom prompts are especially useful when you need:

• Consistent formatting across records
• Clear definitions for what counts as a valid value
• Specific summaries or extracted details

Autofill shows its reasoning alongside the result, so you can quickly verify values and refine instructions over time. This feedback loop makes Autofill more reliable the more you use it.

‍

Run autofill manually or automate data entry

Manually — on demand while you work

Trigger autofill any time — while reviewing a single deal or batch updating a segment of your pipeline. You decide when it runs and which fields it fills.

Automatically — as part of a Streak automation

Add AI Autofill box fields as a step in any Streak automation flow. Set your trigger and autofill runs at the right moment without any manual prompting.

A few examples of how you can automate AI Autofill in Streak:

• Box created → autofill enriches company fields from web research before you open it
• Stage changes → autofill updates a "next steps" or "deal status" field based on the latest emails
• Stage changes → autofill scores ICP fit and updates a qualification column automatically