What it does. Runs nightly across every open Commit and Best-Case deal, comparing what the rep committed against what the engagement actually shows — deals gone quiet, stage-to-stage stalls, missing next steps — and drafts the single question the manager should ask in the forecast review.

The flow. Pull open opportunities → compute staleness, stage-age vs. team median, and the commit-vs-evidence gap → AI ranks risk and writes the inspection question → a ranked digest lands in the forecast doc before the call.

Human checkpoint. The manager runs the questions live; the rep confirms or updates. The model flags risk — it never re-categorizes the forecast or moves a close date.

Metric it moves. Forecast accuracy (call vs. actual) and slippage rate.

What it does. Collapses half a day of research to minutes — org structure, tech stack, triggers, likely pain — and scores each account against your defined ICP so “who do I work first” becomes a ranked list instead of a gut call.

The flow. Enrich the account → AI scores it against your ICP rubric → returns a fit tier, a one-screen brief, and the wedge (the dated, specific reason to reach out now, with a source) → a ranked A/B/C worklist for reps.

Human checkpoint. The rep validates the wedge before outreach; RevOps recalibrates the ICP rubric quarterly from win/loss. The model scores against your definition — it doesn’t get to invent the ICP.

Metric it moves. Reply and meeting rate on A-tier vs. C-tier; win rate by ICP tier.

What it does. Keeps MEDDIC/MEDDPICC alive by pulling the metrics, economic buyer, decision criteria, and champion straight out of call recordings — as a draft the rep confirms — so qualification stops being a form nobody fills in.

The flow. After each call, AI reads the transcript → extracts each MEDDPICC element with the supporting quote → flags the biggest gap (say, an unidentified economic buyer) and drafts the next step to close it → proposes CRM updates for the rep to confirm.

Human checkpoint. The rep confirms before fields commit. Every element must cite a transcript quote; “not established” beats a confident guess.

Metric it moves. Share of Commit-stage deals with complete qualification — and the win rate that tracks it.

What it does. Closes the leak between a conversation and the recorded next action — the place momentum quietly dies — by turning a call into a clean summary, updated CRM fields, and a drafted follow-up in the rep’s own voice.

The flow. Transcript ready → AI writes a five-bullet summary, structured field updates (next step, contacts met, a proposed close-date sanity check), and a follow-up email → all staged for the rep.

Human checkpoint. The rep approves the fields and sends the email. Nothing auto-sends; close-date changes are proposed, never committed.

Metric it moves. CRM completeness, time-to-follow-up, and selling hours recovered per rep per week.

What it does. Points AI the right way — deep research into a single account to write one genuinely relevant message — so reply rates rise instead of the domain burning. Specificity is the point; volume is the trap.

The flow. Take the research brief and wedge from workflow 02 → AI drafts one message per contact, grounded in a single verifiable hook and tied to one business outcome → staged in the sequencer as a manual send.

Human checkpoint. The rep reviews and sends. Every message must pass one test: would a smart buyer believe it was written for them? One cited hook, no fabricated personalization, volume capped.

Metric it moves. Reply and positive-reply rate and meetings booked — with unsubscribe/spam rate as the guardrail.

What it does. Flags the patterns that precede a slipped deal or a churned account — engagement dropping, a champion going dark, a stalled paper process, usage decline — early, while there’s still time to act.

The flow. A weekly sweep across open pipeline and the renewal book → AI scores each record Green/Yellow/Red with the evidence → names the most likely failure mode and the single best save play, with an owner → digest to the AE and CS manager.

Human checkpoint. The AE or CSM runs the play. AI flags; humans act. Thresholds are tuned so Red means Red.

Metric it moves. Slippage caught early, gross and net revenue retention, and win rate on flagged-and-actioned deals.