Agentic Arrow — Case Studies

ES CASE 01 English Stone · Hardscape & masonry, Minnesota

The estimator that prices like the owner.

A rep walks the property, takes measurements, talks through the job. The estimator takes it from there — picks the products, calculates quantities and labor, applies the company’s real margin rules, and writes the branded proposal as HTML and PDF. Estimates that used to eat a rep’s evening come back in minutes, priced with the owner’s own discipline.

Field notes inThe owner’s rulesHTML + PDF out

agentic-arrow: ~/deployed/es-estimator ✓ running

View the walkthrough →

Capabilities

Product selection from the company’s own catalog and pricing rules
Quantities and labor calculated from field measurements
Margin protection applied to every line, every time
Branded proposal — HTML + PDF — generated and filed
Runs from the rep’s phone, same-day turnaround

Live · In production

ES CASE 02 English Stone · Same client, second system

Live nursery inventory, matched to the jobsite.

Crews used to design plantings around what the nursery probably had. This one reads live availability from two regional wholesale suppliers and matches it to the jobsite — sun, exposure, mature size — then hands the crew a pick list they can buy against the same day. Stock first, design against truth.

2 suppliers, read liveSite matchCrew pick list

agentic-arrow: ~/deployed/plantings ✓ running

View the walkthrough →

Capabilities

Live stock pulled from two wholesale nursery suppliers
Matched to site conditions — sun, exposure, mature size
Crew-ready pick list with quantities
Refreshes as inventory moves — no stale catalogs

Live · In the crews’ hands

PCA CASE 03 PCA · National trade association — painting contractors

Four platforms. One member profile.

Membership lived in one system, course streaming in another, email in a third, events in a fourth — four honest platforms that had never met. The build joins them into a single profile per member: who they are, what they watch, what they open, where they show up. Then it turns every event into an action list — who’s coming, who went quiet, who’s ready for more.

4 platforms1 profileEvent actions

agentic-arrow: ~/deployed/pca-members ✓ piloting

View the walkthrough →

Capabilities

Membership, streaming, email, and event platforms unified per member
Engagement history assembled from real behavior, not survey answers
Event actions — who to call, who to re-engage, who to invite deeper
Segments that rebuild themselves as the data moves

Pilot · Built on live data

AH CASE 04 Aleph Holdings · Home-services holding company

Every rep coached every day, by the company’s own playbook.

Reps record their field appointments. Overnight, the pipeline pulls every recording, transcribes it, matches it to the calendar, scores it against the company’s own playbook, and publishes a branded coaching report to the rep’s phone before the next morning. Nobody runs it. It runs.

Sales CoachGAM CoachL10 CoachCanvass Coach

agentic-arrow: ~/deployed/aleph-ai ✓ nightly

View the walkthrough →

The Suite

Sales Coach — every recorded appointment scored against the eight-section sales model
GAM Coach — manager 1:1s graded against the goal-alignment rubric
L10 Coach — leadership meetings scored on facilitation, nine sections
Canvass Coach — door-to-door shifts coached on the company’s knock framework
Nightly: pull → transcribe → match → score → publish. Zero human steps.

Live · In production · Nightly

Case studies. All four running right now.

The estimator that prices like the owner.

Live nursery inventory, matched to the jobsite.

Four platforms. One member profile.

Every rep coached every day, by the company’s own playbook.

Got one workflow that's bleeding hours? Let's aim there first.