PMI Suite

01 Overview

PMI Suite is a portfolio of Philip Morris International Retail Planning and conversational-AI projects: a market-intelligence SaaS ('Retail Planning' / PMI MSE) with an Azure Maps geo-drilldown dashboard, a GenAI/RPA tier (rpi-genai, rpai-capacity-planning) that turns natural-language requests into Snowflake analytics and automated distributor-target calculations, and the 'InsightSpeak' conversational chatbot built on LangGraph agents with AWS Lambda backends and React chat UIs.

Scope of delivery

Built a portfolio spanning a React + Azure Maps retail-planning dashboard (pmi_mse, rpi-rpa-ui), a FastAPI GenAI platform (rpi-genai) with LangGraph orchestrator/SQL/distributor agents over Snowflake, and the InsightSpeak chatbot (React UIs + AWS Lambda LangGraph backends with MS Entra auth).

02 Business Challenge

Deliver PMI retail-planning market intelligence and conversational-AI tools: geo-spatial analytics down to point-of-sale, natural-language access to Snowflake data, and automated distributor-target calculations.

• Azure Maps geo-drilldown (Country to Point of Sale) with demographic and benchmarking analytics.
• GenAI SQL agent with validated text-to-SQL against Snowflake and Excel export.
• Distributor agent that allocates volume targets across stores, with human-in-the-loop interrupts.
• InsightSpeak conversational chatbot with Azure AD SSO and agent backends.

03 Our Solution

Built a portfolio spanning a React + Azure Maps retail-planning dashboard (pmi_mse, rpi-rpa-ui), a FastAPI GenAI platform (rpi-genai) with LangGraph orchestrator/SQL/distributor agents over Snowflake, and the InsightSpeak chatbot (React UIs + AWS Lambda LangGraph backends with MS Entra auth).

Engineering approach

PMI Suite is a portfolio of Philip Morris International Retail Planning and conversational-AI projects: a market-intelligence SaaS ('Retail Planning' / PMI MSE) with an Azure Maps geo-drilldown dashboard, a GenAI/RPA tier (rpi-genai, rpai-capacity-planning) that turns natural-language requests into Snowflake analytics and automated distributor-target calculations, and the 'InsightSpeak' conversational chatbot built on LangGraph agents with AWS Lambda backends and React chat UIs.

04 Design Thinking

05 Technical Architecture

06 Technology Stack

07 Development Timeline

Delivered through structured phases - discovery, design, build, integration, and launch - with iterative releases and ongoing enhancements across a retail planning + genai + chatbot delivery.

08 UI Screens / Key Features Showcase

• Azure Maps geo-drilldown (Country to Point of Sale) with demographic and benchmarking analytics.
• GenAI SQL agent with validated text-to-SQL against Snowflake and Excel export.
• Distributor agent that allocates volume targets across stores, with human-in-the-loop interrupts.
• InsightSpeak conversational chatbot with Azure AD SSO and agent backends.

09 Before vs After

10 Performance Metrics

11 Business Outcomes

12 Testimonial & Connect