Marq - AI CRM Automation: Closed-Lost Deal Revival Agent

Marq's sales team was sitting on hundreds of closed-lost deals with no way to know which ones were worth a second look. I built an AI agent that automatically evaluates old deals at 60, 180, and 360-day intervals, researches the company for revival signals, drafts personalized outreach, and notifies the assigned rep via Slack - replacing an entirely manual process that was inconsistent at best and skipped entirely at worst.

Challenge

Re-engagement decisions at Marq were ad hoc. Reps would occasionally remember to revisit a closed-lost deal, but there was no system, no timing, and no consistency. Revenue was being left on the table.

The solution also needed guardrails. Not every old deal should be resurfaced - some contacts had opted out, some companies already had an active deal, and certain loss reasons (reseller accounts, automation errors) should never trigger outreach. I designed a 5-rule suppression system with input from the VP of Sales and VP of Operations to make sure the agent only surfaces deals worth pursuing.

Approach

Before building, I evaluated 5 deep research APIs (Perplexity, Exa, OpenAI, Gemini, Tavily) and recommended a tiered approach: Perplexity Sonar for MVP, Exa + Gemini for production scale.

The pipeline I built in n8n works like this: when a deal hits 60, 180, or 360 days in closed-lost, a HubSpot webhook fires. The agent runs suppression checks first, then gathers deal context and CRM history, enriches it with Marq's internal knowledge base (product docs pulled from Google Drive via RAG), runs deep company research through Perplexity to surface rebrands, leadership changes, or funding rounds, and sends everything to Gemini for a revival assessment.

If the deal qualifies, the agent drafts a personalized outreach email grounded in Marq's product positioning, creates a new revival deal in HubSpot, and sends the assigned rep a Slack DM with the full context. If it doesn't qualify, it posts the reasoning to a monitoring channel. A separate error handler workflow catches failures and surfaces them via Slack so nothing disappears silently.

Results

• Eliminated manual deal review for re-engagement decisions across the entire closed-lost pipeline
• 5-layer suppression logic prevents inappropriate outreach, approved by VP Sales and VP Operations
• Personalized outreach drafted automatically with Marq product context, ready for rep review before sending
• Reps notified directly via Slack DM when a revival opportunity is created
• Full pipeline validated end-to-end with real closed-lost deals across qualified and disqualified branches
• Pending VP-level sign-off before production activation

Tools Used

n8n, HubSpot API v3, Gemini 3 Flash (via Google AI Studio), Perplexity Sonar Deep Research, OpenRouter, Google Drive API, Slack API