# P13 — Compatibility Matrix: Markets, Telephony, Languages, Platforms

**Updated:** 30 August 2026 · Web-verified 30 Aug 2026 where noted; **[test]** = must be proven by our own calls in P1/P2 before being sold. This file is the honesty ledger — sales may only promise what a row here marks ✅.

---

## 1. Market / telephony matrix

| Market | Local number via platform | Regulatory reality (verified) | **Our serving pattern** | Status |
|---|---|---|---|---|
| **US** | ✅ Easy — Twilio-class local numbers ~$1.15/mo (+ platform $2/mo), inbound ~$0.0085/min, instant API purchase | A2P/robocall rules target *outbound*; our inbound receptionist is low-friction. Two-party-state disclosure handled in greeting (`SECURITY.md`) | Conditional forwarding from client's existing line → platform number | ✅ Ready P1 |
| **Bangladesh** | ⚠️ Constrained — Twilio lists BD under documentation-required regimes; BTRC tightly controls VoIP termination; practical availability poor | Local VoIP number for a BD SMB is not a pilot-friendly path; forwarding a GP/Robi number to an international platform number incurs international charges on the client | **WhatsApp Business Calling** (call button; inbound calls free per Meta) as primary; optional forwarding only if the client accepts charges **[test in P3 with real BD pilot]** | ⚠️ WhatsApp-first |
| **UAE** | ❌ No — Twilio sells no UAE local numbers; UAE blocks/penalizes VoIP-style usage on its numbers (TDRA regime); imported numbers get blocked | Additionally all-party consent recording law (`SECURITY.md` §2) | WhatsApp Business Calling via partner-fronted deployment; no recordings | ⚠️ Partner + WhatsApp only |
| **KSA** | ⚠️ Twilio offers KSA numbers with heavy documentation requirements (regulatory bundles) | Compliance doable but slow; partner entity likely needed for the bundle | Post-P8 only, partner-led | ⬜ Not in pilot |
| Other Gulf (QA/BH/KW/OM) | Unresearched | — | — | ⬜ Gap (finding: research only when a partner brings a live prospect) |

**The strategic read, plainly:** the US is the only friction-free phone market — which is why the LLC channel is the growth path. BD/Gulf are viable primarily as **WhatsApp-voice markets**, which conveniently matches how those customers already call businesses.

## 2. Channel matrix

| Channel | US | BD | Gulf | Notes |
|---|---|---|---|---|
| Phone (forwarded) | ✅ | ⚠️ client-pays-forwarding | ❌ | |
| WhatsApp Business Calling | Possible, low demand | ✅ primary | ✅ primary | Inbound calls free via Meta Calling API **[verify current Meta pricing at first BD pilot]**; requires WhatsApp Business Platform onboarding per client |
| WhatsApp text follow-ups (confirmations, summaries) | ⚠️ SMS preferred | ✅ | ✅ | Template-message costs apply — pennies at pilot volume |
| SMS confirmations | ✅ | ⚠️ | ⚠️ | US-normal; BD/Gulf prefer WhatsApp |

## 3. Language readiness (honest, D10 — the column that gates sales)

| Language | STT quality | TTS quality | Verdict today | Gate to "sold" |
|---|---|---|---|---|
| **English** | ✅ Excellent, all platforms | ✅ Excellent | **LIVE** — sellable now | Already met by platform maturity; still passes per-client QA |
| **Arabic (Gulf)** | ⚠️ Good on major engines; dialect variance real | ⚠️ Good voices exist | **EXPERIMENTAL** — plausible, unproven by us | 5 scripted calls (C-06) + full 20-call script in Arabic before any Gulf client is promised Arabic |
| **Bangla** | ⚠️ Supported by major STT engines but conversational-latency, phone-audio Bangla is exactly where engines are weakest; real-world accuracy unproven | ⚠️ Limited natural voices; premium engines' Bangla young | **EXPERIMENTAL — likely the weakest link.** Do not sell. BD pilots start English or bilingual-greeting ("Bangla speakers, we'll message you back on WhatsApp") | Same C-06 gate; if it fails, that is a published finding, not a hidden one — BD voice waits for the engines, BD WhatsApp-text flows can carry meanwhile |
| Hindi/Urdu (Gulf expat callers) | ⚠️ Decent engine support | ⚠️ | ⬜ Untested | Only on partner demand |

## 4. Platform / integration matrix

| Capability | Retell (primary) | Vapi (fallback) | Notes |
|---|---|---|---|
| Per-minute cost (budget stack) | ~$0.09–0.11 **[verify at signup]** | ~$0.08–0.10 BYO keys | `PRICING.md` §2 |
| Native calendar/booking tools | ✅ Cal.com-pattern | ✅ Google Calendar + Cal.com examples | `API_CONTRACT.md` §2 |
| Numbers | $2/mo bundled | Via Twilio/BYO | |
| WhatsApp calling path | Via SIP/partner integrations **[test]** | Via SIP/Meta Calling API integrations **[test]** | The BD/Gulf pattern rides on this — P1 must confirm the practical route on the chosen platform before BD pilot is scheduled |
| Dashboards for our QA (latency, transcripts, sentiment) | ✅ strong | ✅ adequate | Part of the D1 rationale |
| White-label | ❌ native (wrappers exist) | ❌ native | Irrelevant until ~15 clients (D3) |
| Config portability | Prompts/flows are ours in the template repo | same | Re-platforming ≈ 4–6h/client (`API_CONTRACT.md` §6) |

## 5. Client-side systems

| Client system | Booking | Lead delivery |
|---|---|---|
| Google Calendar | ✅ native tools | — |
| Cal.com (we provision, O5) | ✅ | — |
| Calendly | ⚠️ via Cal.com/middleware pattern 2 | — |
| Our own CRM builds | — | ✅ trivial webhook |
| Any CRM with inbound webhook | — | ✅ standard payload |
| No systems at all | ✅ we provision Cal.com | ✅ Google Sheet + WhatsApp alerts |

Every ⚠️/⬜ that gets tested moves to ✅/❌ **with a date and evidence link** — this file is updated as part of the task that produced the evidence, per the working agreement.