One infrastructure. Configurable AI agents for B2B distribution. Read before any proposal or pricing conversation.
A platform where deploying a new workflow for a new client means configuring, not coding. Each module is a pre-built AI agent that connects to the sources the client already uses — email, WhatsApp, files, ERP. The goal is the heylua.ai model: workflow templates per use case, customized per client. A new client doesn't need an ERP to start. They can start with their inbox and a spreadsheet. Platform fit % tracks how close each module is to that vision.
Modules that share infrastructure. When a client already has one module in a cluster, the next one costs a fraction to add — connectors, data layer, and routing logic are already built.
AI reads incoming discount requests (email + Excel attachments), validates against ERP pricing rules and margin logic, and routes to manager with a 1-click Approve/Reject interface. Verde = auto-approve. Ámbar = manager review. Rojo = auto-reject.
| Client | Stage | Phase |
|---|---|---|
| Grupo Lamosa Peru | PROPOSAL: SENT | 30–40 discount emails/day from sales team |
AI receives inbound requests via WhatsApp, parses the distributor's product codes, pulls matching items from the catalogue, and auto-generates a formatted proposal. Complex or ambiguous cases: system pre-fills key fields for back-office to review before sending.
| Client | Stage | Phase |
|---|---|---|
| Aronlight | PRE-PROP: IN PROGRESS | Back-office generates/sends proposals manually for distributor requests |
AI matches customer technical requirements (wattage, finish, size, application) to the right SKUs from a large catalog. Eliminates manual lookup by engineering team. Outputs a ranked shortlist with confidence scores.
WhatsApp chatbot captures inbound leads, qualifies them through a conversation flow (vehicle type, service needed, availability), and pushes structured lead data to CRM. Eliminates lost leads from untracked WhatsApp conversations.
| Client | Stage | Phase |
|---|---|---|
| AutoalDía | PRE-PROP: IN PROGRESS | 300+ leads/month via WhatsApp; no CRM tracking |
Replaces fully manual returns handling (WhatsApp + Excel + phone) with a structured digital flow. Customer submits return via a simple interface. Back-office processes through a defined workflow. All stock movements tracked automatically.
| Client | Stage | Phase |
|---|---|---|
| Aronlight | PRE-PROP: IN PROGRESS | Fully manual (Excel + WhatsApp + phone). Person responsible just left. Urgent gap. |
AI reads incoming orders sent via WhatsApp to sales executives, parses the order details (SKUs, quantities, client), and pre-fills the SAP entry for the ADV team to confirm with one click. Eliminates the manual copy-paste step between WhatsApp and SAP for every order.
| Client | Stage | Phase |
|---|---|---|
| Grupo Lamosa Peru | PROPOSAL: IN PROGRESS | WF2 candidate. ADV team manually enters all WhatsApp orders into SAP. |
Proactive AI pricing recommendations per customer segment, product line, and competitive context. Complements A1 (reactive approval) by answering a different question: not "should I approve this discount?" but "what should this customer's price be in the first place?" Helps sales reps quote faster and protect margin by default.
| Client | Stage | Notes |
|---|---|---|
| No active client. Best-fit candidates: Lamosa (after A1 live) · Aronlight (natural extension of A2) | ||
Replaces Excel-based raw material management with AI-assisted ordering rules. Calculates reorder points, flags overstock and stockout risk, tracks waste (merma), and surfaces procurement recommendations.
| Client | Stage | Phase |
|---|---|---|
| MAGG | PROPOSAL: SENT | 17K raw material SKUs in Excel; NetSuite goes live April 2026 |
Digitizes machine manuals and maintenance history, builds a preventive maintenance schedule, and sends alerts before failures happen. Moves from reactive (fix when broken) to preventive (schedule before breaking).
| Client | Stage | Phase |
|---|---|---|
| MAGG | PROPOSAL: SENT | All knowledge in technicians' heads; fully reactive; no system |
| Eurostar | PRE-PROP: SENT | 200+ machines; 12-month technician onboarding; knowledge siloed |
Replaces manually-fed Google Sheets with a live dashboard showing SKU performance, stockout signals, and negative sales (lost orders not captured). Semáforo logic: Verde = healthy · Ámbar = watch · Rojo = urgent.
| Client | Stage | Phase |
|---|---|---|
| MAGG | PROPOSAL: SENT | Entry point for Track B; replaces daily manual Google Sheets |
Structured tracking of after-sales service cases. Cases opened via email or WhatsApp, assigned to technicians, tracked through resolution. Director gets real-time visibility. Eliminates cases lost in WhatsApp threads.
| Client | Stage | Phase |
|---|---|---|
| Eurostar | PRE-PROP: SENT | Manual case tracking; director has zero visibility |
| ERP | First connector build | Reuse per module |
|---|---|---|
| SAP B1 / SAP Hana | 3–4 weeks | ~1 week |
| Odoo | 2–3 weeks | 3–5 days |
| NetSuite | 2–3 weeks | 3–5 days |
| Salesforce | 1–2 weeks | 2–3 days |
| Impulsa CRM | 2–4 weeks (custom) | ~1 week |
The moat: connector built once per client. Every new module = days of integration, not weeks. This is what makes renewals easy — switching cost grows with each module added.