The platform, explained honestly.
Plain-English documentation of every agent, pipeline, and moving part โ written from the code and verified against the running system, never from memory. If a page says something works, it was tested the day the page was stamped.
flowchart LR
CUST["๐ค Customer
WhatsApp"] <--> GW["๐จ whatsappGateway
every message, one door"]
GW <--> MAI["๐ค marriageAI
replies to every message"]
OAI["๐ฃ outreachAI
reaches out on its own"] --> GW
MM["๐ matchmakingAI"] --> MD["proposal delivery"] --> GW
WEB["๐ marriage.pk website"] --> API["โ๏ธ coreAPI
api.marriage.pk"]
DASH["๐ฅ๏ธ Staff Dashboard"] --> API
DASH --> INBOX["๐ Agent Inbox
human takeover"]
MAI -.-> INBOX
LEARN["๐ก learning pipelines"] -.-> MAI
classDef hub fill:#e8eef4,stroke:#4e5d6b,color:#182420
classDef agent fill:#e7f2ec,stroke:#166b4e,color:#182420
class GW,API hub
class MAI,OAI,MM agent
The whole system on one map. Every arrow is a real, verified path โ nothing aspirational.
Browse by area
marriageAI
The AI that answers every customer on WhatsApp โ how a single message becomes a reply, which brain handles it, and what it knows about you.
4 pages ยท the context builder, Question Bank and Corrections
The proactive agentoutreachAI
The AI that decides โ like a human would โ whether, when, and what to message customers who haven't written to us.
1 page
Shared by both agentsAgent memory
What the AI remembers about a customer between conversations โ how it is written, who is allowed to see it, and where staff can correct it.
1 page ยท tested live 2026-08-08
Getting smarterLearning pipelines
The six feeds that teach the agents from real conversations โ which ones genuinely work, and what each one does.
2 pages ยท live status board + how scoring works
The front doorwhatsappGateway
Every WhatsApp message, in and out, passes one door โ recording, safety checks, and the free testing band that can never touch a real customer.
1 page