Lightfield
AI-native CRM that builds customer memory and runs follow-up work for you
Lightfield is the high-leverage choice for founders and lean revenue teams who need to replace manual CRM admin with AI-driven customer memory and follow-up automation.
Why we love it
- Builds CRM from inbox, calls, and meetings in minutes
- Meeting prep and follow-ups run from real customer context
- Code execution enables battle cards and pipeline analysis
- REST API and MCP connectors improve stack integration
Things to know
- Still early, and user feedback shows occasional bad auto-created records
- Advanced permissioning sits on the higher Pro tier
- Best for startup GTM, not classic enterprise field sales
About
Executive Summary: Lightfield is an AI-native CRM for founders and early-stage revenue teams. It captures emails, calls, and meetings, builds a live customer memory, and lets agents update records, answer questions, and run follow-up work automatically. It is built for teams that want less CRM hygiene and more execution.
Lightfield says you can connect your inbox, upload a spreadsheet or CSV, and rebuild your CRM in less than five minutes. The product added code execution on Feb 20, 2026, opened its REST API public beta in March 2026, and launched MCP connectors for Notion, Linear, and Granola on Feb 6, 2026. Lightfield offers a Freemium plan, with paid tiers starting at $59 per user per month. It is Less expensive than average for this category.
What makes it stand out is not just auto-capture. It stores customer interactions as continuous context, supports unlimited agent queries and actions on paid plans, and can push data outward through workflow HTTP blocks. For founder-led sales and lean GTM teams, that shifts CRM from passive database to active operating system.
Key Features
- ✓Capture emails, calls, and meetings to remove manual CRM logging
- ✓Answer revenue questions from continuous customer memory instead of scattered notes
- ✓Draft follow-ups and artifacts with code execution against CRM context
- ✓Trigger workflows and HTTP actions to automate downstream GTM ops
- ✓Connect Notion, Linear, and Granola through MCP for richer context sync
- ✓Evolve a configurable data model without upfront CRM schema work
Product Comparison
| Dimension | Lightfield | Attio |
|---|---|---|
| Core use case | AI-native customer memory plus execution layer | Composable startup CRM with strong manual control |
| Time to first value | Connect inbox and CSV, rebuild in under 5 minutes | More setup effort before full value shows up |
| Data capture model | Continuous context from emails, meetings, and calls | Structured CRM objects with customization focus |
| Automation depth | Agent queries, workflows, code execution, follow-ups | Automation is useful, but less agent-centric |
| Integration surface | REST API beta, MCP for Notion, Linear, Granola, HTTP out | Strong composability, but less AI-execution emphasis |
| Best-fit buyer | Founders and lean GTM teams who hate CRM admin | Ops-heavy teams wanting a polished modern CRM foundation |
Frequently Asked Questions
The core difference is automation depth. While Attio excels at composable CRM structure and a more mature setup, Lightfield has an absolute advantage in self-building CRM, sub-5-minute onboarding, meeting prep, follow-ups, and agent-driven analysis from continuous customer memory.
It is promising, but still early. Public feedback praises speed and automation, yet some users flagged AI-created bad records or nonexistent companies. The practical workaround is to start with limited imports, approval-based workflows, and human review before bulk updates.
Yes. Startup includes a free trial, then starts at $59 per user monthly, while Pro is $149 per user and billed annually. Pro adds higher limits, advanced permissioning, dedicated onboarding, and a dedicated customer success manager.
It fits well as both memory layer and action layer. Lightfield connects email and calendar data, added MCP connectors for Notion, Linear, and Granola, opened a REST API public beta, and supports HTTP-out workflow steps for downstream automation.
Partially, but teams should verify before rollout. Public materials show configurable data models and advanced permissioning on Pro, yet they do not fully spell out enterprise privacy details on the pages reviewed. Sensitive teams should request security documentation during pilot.
Yes. Since Feb 20, 2026, Lightfield has supported code execution, letting the agent write and run Python against CRM memory. That makes battle cards, pipeline analysis, account plans, and structured artifacts much more practical than standard call-note tools.