AI sales automation: lead capture, research, follow-ups and proposals
Agents handle the research, scheduling and first drafts around each sales step, a person approves anything customer-facing, and the CRM stays up to date.
The problem
Sellers and estimators spend hours on work around the sale: typing in business cards, researching accounts before calls, finding meeting times, writing agendas and follow-ups, and updating the CRM afterwards. When it gets skipped, deals sit with stale amounts and next steps, and customer questions go unanswered.
What we build
We automate each step around the sale with a workflow tool, the CRM and an AI agent. Inputs are captured once, the CRM is searched before anything is created, the agent researches and drafts, the calendar is checked before booking, and a person approves every customer-facing message before it becomes a draft. Our Zapier walkthrough shows five of these flows running on HubSpot, Gmail, Google Calendar and OpenAI; proposal drafting extends the same pattern to your templates.
How it works
- 1
Capture the lead once
A business card photo posted in Slack or a form submission starts the flow. OCR reads cards, and the CRM is searched for the company and contact before either is created, so duplicates stay out.
- 2
Research and prepare
An agent researches the person and account and writes a CRM note. For a discovery or demo request it checks the calendar, books the meeting with a generated agenda, or emails the account executive to pick another time.
- 3
Follow up and update the deal
After the call, the recording is transcribed and the agent drafts the follow-up and, where you use them, a proposal from your template. Deal amount, close date and next step are cleaned and written back to the CRM.
- 4
A person approves what goes out
Customer-facing drafts go to a person who approves or sends them back with feedback. Approved emails are saved as drafts in Gmail rather than sent automatically.
- HubSpot
- Zapier, n8n or Power Automate
- Gmail and Google Calendar
- Google Forms and Google Sheets
- Slack
- OpenAI or Claude
- OCR (Nanonets in the walkthrough)
The value it creates
Time saved
Measured by timing the manual research, scheduling and CRM updates around a sample of recent deals against the same steps with the automation.
Standardization
Every lead gets the same lookup, research, agenda and follow-up, regardless of who owns it or how busy the week is.
Visibility
Deal amounts, close dates and next steps are written back after each call, so the pipeline reflects what actually happened.
Collective intelligence
Research notes and call summaries land in the CRM where the whole team can see them, instead of one seller's inbox.
Proof
Related use cases
- Built and shown
AI assistant for project drawings, specs and documents
Ask questions of a project's drawings, specs and meeting notes and get answers that cite the file and passage they came from.
- Built and shown
AI schedule builder for CPM schedules with evidence and XER export
Specialized agents draft a CPM schedule from project documents and client answers, show the evidence behind each activity, and export XER for the scheduler to review.
- Built and shown
MCP servers for Procore, Primavera P6 and other construction systems
An MCP server exposes a construction system's API as tools an AI agent can call, with scoped permissions and a log of every call.
Frequently asked questions
Will the AI email customers on its own?
No. In the flows we build, a person approves every customer-facing message, and approved emails are saved as drafts rather than sent. In the walkthrough a draft is rejected as not technical enough and goes back through the agent before a new version is produced.
Does it work with a CRM other than HubSpot?
The walkthrough uses HubSpot, but the pattern works with any CRM that has an API, such as Salesforce. The workflow tool can be Zapier, n8n or Power Automate depending on what you already use.
Can it write proposals or estimates?
It can draft a proposal from your template, the call transcript and CRM details for a person to edit. The walkthrough shows research, agendas, follow-ups and deal updates; proposal drafting is built to your templates and pricing rules, and pricing is always set by a person.
Is this only for construction companies?
No. The walkthrough comes from a solutions architect's sales workflow, not a construction one. The same pattern fits contractors pursuing work, construction technology vendors and service businesses beyond construction.
Next step
Which report or workflow would you like to improve?
Tell us what your team does today, which systems are involved, and what you want to change. We'll discuss whether there is a practical fit.
Prefer email? charley@buildflows.ai