After a strong sales conversation, the next step is often turning call notes into a clear, professional proposal. This can involve reviewing pain points, objections, decision criteria, budget signals, next steps, and buying intent before preparing a document that accurately reflects the opportunity.
This use case explores how a connected AI automation workflow can turn structured sales-call analysis into a proposal document, distribute it to the right team members, log it for future reference, and automatically enrol the prospect into a follow-up sequence.
The Problem
After a sales call goes well, someone still needs to turn scattered notes, pain points, objections, budget signals, and next steps into a polished, client-ready proposal.
This process can take time, be delayed by competing priorities, and vary in quality depending on who prepares the document. When proposal generation is handled manually, sales teams may struggle to maintain a consistent process from call analysis through to proposal delivery and follow-up.
The Solution
The Proposal Generator Agent receives structured call-analysis output from an upstream workflow, combines it with the sending company’s information, and uses AI to create a full contract-style Proposal or Statement of Work as clean HTML.
The workflow converts the proposal into a PDF, distributes and logs the document, then automatically sends the prospect into a sales follow-up sequence. It also tracks daily and on-demand proposal-generation activity.
How the Agent Works
- Receive sales-call insights. The workflow receives structured information from a completed sales conversation, including buying intent, pain points, objections, decision criteria, budget signals, and next steps.
- Combine client and company information. The agent brings together call-analysis output with the sending company’s details to create the basis for a complete proposal.
- Generate a client-ready proposal. AI produces a structured proposal or Statement of Work using the available information, without inventing missing details.
- Convert, distribute, and log the document. The proposal is converted into a PDF, shared with the team, and saved for tracking and future editing.
- Trigger follow-up and track activity. Once the proposal is created, the workflow automatically enrols the prospect into a follow-up process and records proposal-generation metrics.
Technical Workflow
1. Intake and parallel data gathering
A webhook receives a structured payload from an upstream call-analysis workflow: interest level, buying intent, a call summary, pain points with quoted evidence, objections, decision criteria, a budget range, next steps, and a BANT breakdown.
Two things happen in parallel: a Google Sheets lookup pulls the sending company’s own details, and an Edit Fields node preserves the raw webhook body. Both streams are recombined in a Merge node.
2. AI-drafted proposal (the core step)
The merged data is handed to an LLM agent (GPT-4.1-mini) running an unusually detailed system prompt. It explicitly distinguishes provider fields from client fields, and produces a complete, self-contained, PDF-ready HTML document with 14 required sections.
The prompt is strict about never inventing facts: anything not in the input must be explicitly marked “Not specified in the provided data.”
3. Parsing and PDF conversion
A code node parses the agent’s structured JSON output (document_draft + recommended_next_step) and extracts the raw HTML.
That HTML is POSTed to the PDF.co API to render it into an actual PDF file, which is then downloaded back into the workflow.
4. Distribution and logging
The generated PDF is sent to Slack as a file upload. There is also a disabled Google Drive upload/download pair.
The full proposal HTML is logged to a “Proposal Agent” Google Sheet, keyed by a team/user ID, explicitly designed to support a future correction/editing workflow.
5. Auto-triggering follow-up
Once the proposal is saved, an HTTP request immediately fires to the Sales Follow-up Agent webhook, passing company name, today’s date as the proposal-sent date, and the client’s email.
This automatically enrols the new proposal into the follow-up cadence without anyone manually kicking that off.
6. On-demand proposal metrics
A second webhook accepts a date range and, via a “Proposal Date” tracking sheet, calculates total proposals generated and a daily breakdown for that window.
An LLM (GPT-4.1) turns this into a 1–3 sentence executive summary highlighting totals and any notable spikes or drops.
7. Scheduled daily metrics
Every night at 11 PM, a separate scheduled job counts how many proposals were generated that day from the “Proposal Date” sheet and appends the tally to a “Proposal Metrics” sheet, creating a standing daily log without anyone requesting it.
Technology and Integrations
Built with: n8n, OpenAI GPT-4.1-mini and GPT-4.1, PDF.co, Google Sheets, Slack, Google Drive, webhooks, and connected sales workflows.
Outcome
The Proposal Generator Agent creates a more structured path from sales-call analysis to proposal delivery.
It turns qualified sales-call insights into a client-ready proposal document, distributes it to the team, logs it for future reference, and automatically starts the follow-up process.
The workflow also provides daily and on-demand visibility into proposal-generation activity, helping sales teams track how many proposals are being created over time.
Explore Custom AI Automation for Your Business
The Proposal Generator Agent is one example of how AI automation can support a more structured sales process.
Divverse Labs designs and builds custom AI agents, automation workflows, internal tools, and connected systems for a wide range of business processes.
From sales, marketing, recruitment, customer support, and reporting to finance, operations, and internal team workflows, each solution is designed around the way your business works.
Request a Quote for a Custom AI Automation Solution to get started.