HyAlly
Operations Intelligence in flight

A premium consumer brand's B2B sales team gets an AI assistant that captures everything and forgets nothing

A 5-person global sales team covers airports, distributors, and big-box retail. The HyAlly playbook stood up the data warehouse, replaced the legacy ERP CRM with Odoo, and put a Claude-powered follow-up assistant on every desk in 12 weeks.

Sector

Premium Consumer Brand

Solution

Operations Intelligence

Status

Foundation SOW in flight. Sprint 7 extension running into early Q3.

12 weeks foundation, sales online from week 4

Challenge

  • The company had failed an ERP CRM rollout previously. The sales team would not adopt anything that felt like more admin.
  • Follow-ups fell through the cracks. Five reps covering the entire world. Things missed by days, not hours.
  • Sales pricing happened without seeing data costs or historical context.
  • No way to detect when a B2B customer's order pattern changed by more than a normal amount.

Approach

  • Auto-capture from Outlook and the ERP. No new input fields. The system records what already happens.
  • Odoo CRM replaces the legacy CRM for B2B. Bidirectional sync keeps confirmed orders flowing back.
  • Claude on Bedrock drafts contextual follow-ups. The rep reviews and sends.
  • Escalation rules ping until a logged action dismisses them. The VP gets a daily accountability digest.
  • Reorder model learns each account's cadence and alerts when orders drift by more than 10%.

What shipped

  • AWS foundation, VPC, RDS, EC2 app server, Kestra automation engine, Redis queue for ERP API concurrency.
  • ERP data pipeline, orders, customers, items, GL syncing into a warehouse with raw, staging, warehouse, and reporting schemas.
  • Odoo CRM with full B2B pipeline, territory assignment, Outlook sync via Microsoft Graph, and hot customer alerts.
  • Sales follow-up automation with escalation logic and Claude-drafted email sequences.
  • Claude conversational BI on the warehouse with role-based access.
  • Distributor discovery with contact-enrichment and fit scoring.
  • Reorder model with customer health dashboard.

Outcomes

  • Sales team using Odoo daily inside Sprint 3.
  • Bulk natural language queries on the warehouse from Sprint 4.
  • New prospect leads flowing into the CRM with daily digests from Sprint 5.
  • Proactive reorder alerts live for every active B2B account by Sprint 6.

Stack

  • AWS (single VPC
  • RDS
  • EC2)
  • Kestra automation
  • PostgreSQL warehouse with dbt-style schemas
  • Odoo CRM
  • Microsoft Graph
  • contact enrichment provider
  • Claude on Bedrock
  • Metabase BI
  • ERP integration
  • SES

What's next

Phase 1 alternatives, Sales and Marketing AI Implementation, and Product Launch and Integration Readiness, sit on top of this foundation and are mapped, ready to start when the client signals.

Quote (pending sign-off)

"Placeholder pending sign-off from the E-commerce and Digital Marketing lead."

E-commerce and Digital Marketing, Premium Consumer Brand

More work

Related case studies

Put HyAlly on your own work product.

See it draft, review, or govern in your stack. We will scope a pilot in two weeks.