CRM

AI Agent Triage for Ambiguous Contact Merge Conflicts

A Paperclip agent works through a queue of merge conflicts that rules couldn't resolve, enriches each candidate with web and CRM context.

CategoryCRM
Enginepaperclip
Difficultyadvanced
Triggerschedule
Steps7
Setup~25 min

How it runs

The automated pipeline, trigger to output.

  • TriggerSchedule drains conflict queuePostgreSQLPostgres
  • ActionEnrich pair with web + CRM contextExa
  • LogicAgent reasons and writes recommendation
  • LogicBranch by confidence (auto vs review)
  • ActionRoute uncertain cases to Slack with rationaleSlack
  • ActionExecute approved merges in HubSpot + AttioHubSpotHubSpot
  • OutputAudit decision and evidence to PostgresPostgreSQLPostgres

What it does

Hands the hard dedup cases — the ones deterministic rules punt on — to an agent. For each ambiguous pair it gathers context (company website, CRM activity history, overlapping deals) to judge whether two records are truly the same person, then writes a plain-language recommendation. Confident merges execute automatically; the rest go to a human with the agent's reasoning attached.

When to use it

When your conflict review queue is a recurring time sink and you want an agent to do the investigative legwork so reviewers approve in seconds instead of researching each pair.

How it works

  1. 1A schedule wakes the agent to drain the conflict queue from Postgres.
  2. 2For each pair, the agent enriches with a web lookup and pulls activity from HubSpot and Attio.
  3. 3It reasons over the evidence and writes a merge recommendation with confidence.
  4. 4Branch: high-confidence merges execute; uncertain ones route to Slack with the rationale.
  5. 5Execute approved merges in both CRMs.
  6. 6Record the agent's decision, evidence, and outcome in the Postgres audit log.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect PostgresAny Postgres URL — query, write, migrate.
  2. 2
    Connect HubSpotCRM, deals, marketing, support.
  3. 3
    Connect AttioReal-time CRM with structured data + powerful views.
  4. 4
    Connect ExaNeural search across the web.
  5. 5
    Connect SlackChannels, DMs, threads, mentions.
  6. 6
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  7. 7
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  8. 8
    Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.

Run this workflow in your colony.

14-day trial. No DevOps. No Sales call. Provisioned in under a minute.