The challenge
A mid-tier pharma company with a 240-rep field force across 8 states was losing field intelligence at the source. Reps were supposed to log every HCP visit into a Veeva-style SFA app — visit notes, samples handed over, prescription intent — but compliance was 63%. Notes were generic ("Doctor positive"). Sample reconciliation at month-end was a 4-day finance exercise. Brand managers had no signal on which doctors actually moved the needle.
How we deployed
- Replaced the visit-form with a 60-second voice recording the rep dictates after each chamber visit.
- Built an AI pipeline that transcribes, structures and writes back into the SFA: HCP, brand discussed, sample SKU + qty, sentiment, next-visit recommendation.
- Auto-reconciled samples handed over against the rep stock register every night.
- Surfaced brand-by-doctor sentiment trends in a brand-manager dashboard.
- Flagged visit-frequency anomalies — reps over-visiting low-prescribers, under-visiting top-tier KOLs.
- Routed coaching prompts to ASMs based on the actual content of rep conversations.
What changed
- Each rep reclaimed 11.8 hours of weekly admin time — equivalent to roughly 2 extra HCP visits per day.
- Visit-log compliance climbed from 63% to 96% across the field force.
- Sample reconciliation accuracy hit 99.1%; finance closed the cycle in hours not days.
- Brand managers got first usable signal on which HCPs were actually being moved.
- Two product launches were re-targeted mid-quarter on the back of the new HCP-level data.
"My reps used to lie on the form because the form was painful. Now they speak for one minute and the CRM is honest."
— National Sales Manager · Mid-tier pharma