Back to Projects

Roddy AI-Agent Platform

Production multi-tenant B2B platform for AI agents, serverless workflows, and enterprise integrations

Roddy AI-agent platform

Role

Technical Lead / Software Architect

Timeline

2025 - Present

Technologies

AWS Lambda, API Gateway, DynamoDB, SQS, S3, CloudFront, AWS SAM, TypeScript, React, Python, Django, Odoo, Epicor, WhatsApp, LLMs

Project Overview

Roddy is a production multi-tenant B2B AI-agent platform used by real clients to automate sales and operational workflows across WhatsApp, email, webchat, ERPs, CRMs, and internal systems.

I designed and built the core technical surface of the platform: the AWS serverless backend, client-facing React/TypeScript portal, middleware, authentication, integrations, deployment, documentation, and production configuration.

Architecture

Serverless Backend

AWS SAM backend with 50+ Lambda functions, 30+ DynamoDB tables, 8 SQS queues, API Gateway, S3, CloudFront, Lambda Layers, asynchronous processing, and event-driven workflows.

Client-Facing Portal

React/TypeScript platform portal for managing clients, users, agents, channels, connections, use cases, webhooks, events, notification rules, dashboards, and web chat.

AI-Agent Runtime

Dynamic runtime configuration, reusable AI tools, configurable use cases, multi-channel messaging, tenant isolation, and extensible agent orchestration.

Enterprise Integrations

Integrations with Odoo, Epicor, custom ERP middleware, WhatsApp, email, webchat, Google Calendar, Google Spaces, and client-specific business workflows.

Impact

Roddy supports production usage across approximately 7 B2B clients, processing 100+ contacts per day and several thousand automated messages per month.

For a large Mexico-based office supplies distributor, the platform helped automate lower-touch sales workflows, reduce manual sales workload, and enable 24/7 automated response handling.

What This Demonstrates

  • Technical leadership and end-to-end platform ownership
  • Production AWS serverless architecture
  • Multi-tenant B2B SaaS-style product design
  • AI-agent and LLM tool orchestration
  • ERP/CRM/API integration under real business constraints
  • Client-facing product delivery, documentation, and operations