Abu Asif Khan Chowdhury

Python/Django

• · LangChain/LangGraph

◦ · GPT-4

◦ · Claude

◦ · RAG

◦ · Redis

◦ · Celery

◦ · PostgreSQL

◦ · Railway

◦ · iOS/Swift

Product & Business: Built Donna AI — a unified AI assistant for email, Slack, WhatsApp, and calendar that reduces communication triage from hours to minutes. Designed gamification engine driving user engagement and retention. AI Architecture: Engineered multi-agent system (GPT-4, Claude, LangChain, LangGraph) with RAG pipelines and persistent memory for deep personalization. Built relationship intelligence engine with RFM scoring and DISC personality analysis. Developed custom eval framework for continuous model quality assurance. Engineering: Architected Django/PostgreSQL/Redis/Celery backend with plugin-based adapter pattern supporting Gmail, Outlook, Slack, and WhatsApp integrations. Shipped native iOS app with streaming chat, voice/TTS, and guided onboarding. Managed production stability across Railway environments. Team & Culture: Led team of 4 engineers + 1 UX designer with spec-driven workflow. Built AI-augmented dev pipeline using Claude for code review and automated eval harnesses. Full product lifecycle ownership from ideation to production.

Microsoft Semantic Kernel

• · Azure AI Search

◦ · Azure Foundry

◦ · RAG

◦ · Azure

◦ · CI/CD

◦ · IaC

AI Platform: Architected full-stack AI integration using Semantic Kernel, Azure AI Search, and Azure Foundry. Built RAG pipeline enabling natural language querying across massive data warehouses, transforming weeks of SQL work into conversational insights in seconds. Developed agentic AI systems that autonomously retrieve, synthesize, and contextualize complex data — eliminating manual analysis bottlenecks. SOC2 Compliance: Designed entire Azure cloud platform from ground zero with SOC2 baked in. Led collaboration with Thoropass to achieve SOC2 Type II certification. Built compliance-ready infrastructure that became a competitive differentiator for enterprise deals. DevOps: Provisioned scalable Azure infrastructure with CI/CD pipelines, IaC, and monitoring. Implemented high-availability architecture supporting AI workloads at scale.

C#

• · DotNet

◦ · Javascript

◦ · Typescript

◦ · Microsoft SQL Server

◦ · Microsoft Azure

◦ · Dataverse

◦ · Power Platform

◦ · X++

Asset Management Mobile App: Led end-to-end engineering — requirement gathering, UX brainstorming, performance analysis, sprint planning, and implementation. Released to public preview, very well received by customers. Supply Chain Management (Dual-Write): Made dual-write for SCM more resilient, enabling seamless Dataverse-FnO data flow. Reduced customer pain score from 10 to 1 while improving resilience, performance, and documentation. Dynamics 365 SCM Trade: Reduced daily customer incidents from ~40 to under 10. Led a team of 9 vendors, ensuring highest quality fixes and the lowest work item count in product history. Dynamics 365 Project Service: Designed and implemented task-based billing system and retainer-based contracts using graph-based algorithms. Features received positive customer feedback with minimal issues. Dynamics 365 Marketing: Improved performance by removing redundant middle-layer microservices. Achieved 100% accessibility compliance rating. Enterprise Marketing: Supported hundreds of customers on Dynamics 365 Customer Engagement in-app marketing. Worked closely with customers to develop generic fixes benefiting all users.