Automated coaching for every sales meeting. Every rep gets personalized feedback within minutes. Management gets weekly visibility into team performance, skill gaps, and deal risks — without listening to a single recording.
Managers can't listen to every call. Reps don't get consistent feedback. Coaching gaps compound into lost deals. By the time a pattern is visible, it's too late to fix.
The pipeline runs every 15 minutes. When a Zoom recording lands in HubSpot, it flows through transcription, analysis, and delivery without anyone touching a button.
Detects completed Zoom calls with recordings
Deepgram Nova-3 with speaker diarization
Claude scores against configurable rubrics
Slack DM to rep + weekly rollup to managers
Scores & notes written back to HubSpot
Reps, managers, and the CRM all get the information they need — in the format and cadence that makes sense for each.
Slack message sent to the rep within minutes of each meeting. What went well, what to sharpen, and a specific next action.
Team-wide summary posted to a private Slack channel every Monday. Flagged meetings, rep performance, objection patterns, and deal risks.
Coaching scores and detailed notes written directly to call, contact, and deal records — so CRM data stays current without manual entry.
Each meeting type is scored against different dimensions that reflect what matters at that stage of the sales process. Rubrics are config — editable without changing code.
4 global dimensions shared across all types + call-type-specific dimensions. Scored 1–100 per dimension, weighted average overall.
Every sales meeting is analyzed — not just the ones a manager happens to sit in on. No meetings fall through the cracks.
Reps get coaching in Slack within minutes of the meeting ending — while the conversation is still fresh.
Coaching scores, meeting notes, and deal insights are written directly to HubSpot. CRM stays current automatically.
Objection trends, skill gaps, and deal risks surface in the weekly rollup — before they become lost deals.
The core pipeline is operational. Near-term work focuses on making the coaching data more accessible and actionable through dashboards and on-demand reporting.
HubSpot polling → Deepgram transcription → speaker identification → call type classification. Runs every 15 minutes.
Claude scores each meeting against configurable rubrics. Extracts objections, next steps, and coaching recommendations.
Automated Slack DM to the rep after every scored meeting with personalized feedback and recommended actions.
Scores and coaching notes written to call, contact, and deal records. Deal reconciler handles pre-deal meetings retroactively.
Monday morning Slack summary with team stats, flagged meetings, per-rep highlights, and objection patterns.
Move from JSON files to Postgres/Supabase — prerequisite for dashboards and faster queries across coaching history.
Rep view: personal call history, score trends, dimension heatmap. Manager view: team comparison, cross-rep analysis.
Reps trigger execution gap analysis and deal-specific reports directly from Slack. Rubric configuration via UI instead of JSON.
Accumulated per-rep coaching profiles and cross-team patterns that improve coaching context and recommendations over time.