$ cd ../projects
flydo/README.md
side projectJan 2026 - ongoingfeatured

Flydo

AI Agent SaaS Platform with React Native Mobile App

problem.md

Businesses need AI support agents that answer from company knowledge, work across communication channels, and remain governed by tenant roles, quotas, security, and admin controls.

role.md

Full Stack Developer - built the agent engine, multi-channel integrations, React Native mobile app, tenant controls, admin tooling, and security workflows.

challenges.md
!Connecting one AI agent system to Telegram, Discord, Slack, Email, WhatsApp, web widget, React Native mobile app, and mobile APIs
!Keeping answers grounded with RAG knowledge bases, correction detection, and persistent user memory
!Managing subscription tiers, quota enforcement, feature gating, tenant RBAC, audit logs, and platform admin sessions
!Protecting public-facing AI workflows with rate limiting, IP blocking, XSS sanitization, and honeypot detection
solutions.md
>Built an AI agent engine with Groq LLM, LLaMA 3.3 70B, multi-round function calling, built-in tools, and custom webhooks
>Implemented a RAG pipeline with TF-IDF embeddings, conversation learning, correction detection, and persistent memory
>Added Telegram, Discord, Slack, Email, WhatsApp, web widget, Socket.IO/Web Push notifications, a React Native mobile app, and 18 mobile API endpoints
>Designed multi-tenant RBAC, a 4-tier subscription engine, OTP admin sessions, feature flags, audit logging, and security controls
$ cat description.txt

A multi-tenant SaaS platform for building and deploying AI-powered conversational agents with RAG knowledge bases, multi-channel support, a React Native mobile app, mobile APIs, notifications, and subscription-based access control.

impact.log

Turned AI support setup into a repeatable SaaS workflow where one trained agent can be deployed across web, messaging, and mobile channels with subscription-aware access.

$ git diff --stat
+Groq LLM and LLaMA 3.3 70B agent engine with multi-round function calling
+RAG pipeline with TF-IDF embeddings, correction detection, and persistent memory
+Omnichannel deployment across Telegram, Discord, Slack, Email, WhatsApp, web widget, React Native mobile app, and mobile APIs
+Socket.IO and Web Push notifications
+Expo APK build for the mobile companion app with auth, dashboard, agents, knowledge bases, conversations, tickets, customers, analytics, and settings
+Multi-tenant RBAC, 4-tier subscription engine, admin panel, feature flags, audit logging, and abuse protection
tech-stack.txt
Next.jsNode.jsReact NativeExpoGroq LLMRAGSocket.IORBAC