Sales re-engagement dashboard that surfaces and prioritizes old prospects
Build a sales re-engagement desk: a tool that resurfaces prospects a rep has already worked — people they called or emailed, contacts who re-open old emails or show fresh buying signals, anyone who asked to be contacted later, and deals that went Closed Lost — ranks them by how likely they are to re-engage now, and reminds the rep when it's time to reach out again. Every action a rep takes in the dashboard should mirror back into the CRM so the system of record stays accurate. Triggers A daily scheduled run (cron, early morning) that drives the data pipeline. An always-on dashboard web page the rep opens in the browser. A manual "sync now" trigger so a rep can refresh on demand. Gate the dashboard and its data/action endpoints to signed-in users, and use the authenticated user's identity (the spoof-proof authenticated-email header) to scope what each person sees — never trust client-supplied identity. Pipeline (high level) Pull the rep's book and relevant history from the CRM — open/worked records, Closed-Lost opportunities, account ownership, and any account-team pairing so a rep and their aligned partner share a book. Enrich with conversation-intelligence signals (call activity, follow-up engagement) and third-party intent/buying signals. Pre-filter aggressively before any AI: gate on cheap metadata, only send genuinely useful material (e.g. connected-call transcripts, trimmed) to the model, and let passive signals skip the model entirely — this keeps token cost down. Use an LLM to read call/context history for explicit callback intent and to classify each account's re-engagement intent (e.g. High/Moderate/Low) with a suggested re-engagement date and a short rationale. Cache analysis so nothing is re-processed. Merge everything into one record per person, compute a tiered intent score and a due date, resolve ownership, and store in a durable database, preserving each prospect's status and history across runs so unactioned people carry forward. Optionally push a daily reminder digest to the rep's chat or email. Dashboard A ranked, grouped worklist (overdue / due / upcoming / snoozed) with filters/slicers, ownership and due-date lenses, and per-prospect actions: mark done, snooze, re-open, disqualify, dismiss. Each prospect expands to show full context — the AI's plain-language reasoning, intent score/tier, best time to reach and connect rate, nurture signals, a "why now" evidence timeline, key CRM facts, and an action history. Reads come from a data endpoint; actions post to an action endpoint that writes to the database and mirrors to the CRM, crediting the acting user. Enforce row-level visibility server-side (a rep sees their own book plus their aligned partner's; admins can see all). Storage Use one durable named volume for the prospect database, with exclusive-writer steps for the sync and action paths and read-only mounts for the dashboard/reminder/read paths. Connectors / external services (all swappable) CRM for records and write-back — Salesforce, or your team's CRM. An LLM for intent analysis — Anthropic Claude, or another model provider. Conversation intelligence for call activity — Gong, or your call/meeting tool. Intent/buying-signal data — 6sense, or your intent provider. Reminder delivery — Slack, or your team's chat tool, and/or email. Design each external source to degrade gracefully to a realistic demo dataset when its connector isn't attached, and read base URLs/credentials from the injected connector environment so no secrets live in code. Tools used: Salesforce, Anthropic Claude, Gong, 6sense, Slack, email, SQLite (durable volume storage)
What this prompt builds
This workflow automatically identifies and prioritizes prospects ready for re-engagement by analyzing Salesforce records, Gong call transcripts, and intent signals to surface timing requests and buying signals. It ranks prospects by likelihood to convert, sets calendar reminders, and presents full context in a dashboard so sales reps can reconnect at the right moment. The system mirrors all actions back to the CRM and reduces lost pipeline by ensuring no opportunity slips through the cracks.
The problem
Sales teams lose pipeline when prospects who asked to be contacted later—or who went quiet after a demo, a Closed-Lost deal, or a scheduling conflict—slip through the cracks because reps lack a systematic way to track and re-engage them at the right time. This workflow reads Salesforce activity logs and Gong call transcripts to identify every prospect who requested a future follow-up, surfaces fresh buying signals and intent data, and ranks each contact by their likelihood to re-engage now. It then delivers a prioritized dashboard with full context—AI-generated timing recommendations, past conversation summaries, and a "why now" timeline—and automatically sets calendar reminders so reps know exactly when and why to reach back out.
Solution and impact
Sales reps regain control of their book by working a ranked, actionable list of warm prospects instead of manually searching CRM records or re-listening to calls to remember who to contact next. The workflow removes the friction of tracking follow-up timelines, ensures no opportunity is forgotten, and positions the team to reconnect at the moment a prospect is most likely to buy. Teams see higher pipeline generation, better account coverage, and improved conversion rates because every re-engagement happens with full context and at the optimal time.