Dhaka, Bangladesh · Open to relocate

Shahtaz Rahman

Backend Software Engineer · RAG & LLM Applications

I build backend systems and cost-efficient AI applications to address high-impact business challenges. From scalable architecture and tools integrations to end-to-end product development, I turn product concepts into purposeful software.

View Projects View Resume

3+ YEARS PROFESSIONAL EXPERIENCE
PYTHON · DJANGO · FASTAPI
AI AGENTS · RAG · LLMs
AWS · GCP

01About Me

I bring 3+ years of software engineering experience, with a focus on backend systems and AI-powered applications. Over the last two years, I've worked on RAG architectures, hybrid retrieval, multi-agent systems, and real-time voice agents. I enjoy owning the engineering lifecycle — making architectural decisions, building the application, optimizing its performance, and getting it into production.

Stack

Languages: PythonTypeScriptJavaScriptCC++
Backend: DjangoFastAPINode.jsCeleryGraphQLWebSockets
AI: Google ADKRAGpgvectorPineconeVertex AI
Data: PostgreSQLRedisMongoDBFirestoreKafka
Cloud: AWSGCPDockernginxCI/CD
Frontend: ReactNext.jsRedux ToolkitTailwind CSS

What I do

AI agents & RAG

AI-powered applications built around business data, with custom ingestion pipelines, hybrid retrieval, agentic workflows, tool integrations and real-time voice capabilities.

Backends & Integrations

Scalable Python backends, REST APIs and asynchronous processing systems that connect applications with third-party services, messaging platforms and business tools.

Full-stack Development

End-to-end product development spanning database design, backend architecture, React interfaces and cloud deployment, with a focus on maintainability and performance.

3+ years: Software Engineering Experience
20+: Projects Delivered

See experience

02Projects

AI agents, data-heavy platforms and developer tools: each with its architecture and the reasoning behind it.

AI Chatbot Assistant for Business & Ecommerce

Vegabot

AI chatbot assistant that a business trains on its own knowledge, products and services. It answers by text or real-time voice conversation, recommends products and services, adds items to the cart, takes orders.

Backend & AI systems engineerCASE STUDY

AI-powered travel planning platform

Tourtoise

A multi-agent travel platform that turns a traveler's preferences into grounded, budget-checked, day-by-day itineraries, then guides them through each day of the trip based on where they are.

Backend & agent systems engineerCASE STUDY

Forex broker management & analytics platform

Restro FX

A broker management platform for forex traders: connect a TradeLocker account and get trading data pulled in automatically, with analytics on top.

Full-stack engineerCASE STUDY

Full e-commerce platform

E-commerce Platform

A complete e-commerce system covering products, variants, inventory and the full order workflow, from cart to delivery.

Full-stack engineerCASE STUDY

Other builds

Network MonitorChrome extension · Developer tool

A Chrome extension that captures a page's REST, WebSocket and SSE traffic in an in-page panel for inspection and debugging.

Source Download .zip

03Career

Across professional teams and independent projects, I've worked on backend architecture, AI-powered applications, and software delivery taking ownership of systems from initial design through implementation and deployment.

Software Engineer (L2)

Current

Echologyx Ltd.

Mar 2024 – Present2 yrs 9 mos

Builds AI chatbot products that answer customers and recommend products for ecommerce businesses, across web, social and email.

Multi-agent rebuild, ~3× faster Led R&D to move the chatbot engine from LangChain, OpenAI and Elasticsearch to a multi-agent architecture on Google ADK, cutting replies from ~12s to ~4–5s with caching and context compaction.
Cheaper, better retrieval Benchmarked Vertex AI against Pinecone and moved to Pinecone hybrid search for product recommendations and follow-up answers at lower cost.
One knowledge base, every channel Connected Messenger, Instagram, WhatsApp and email (Amazon SES, Gmail and Outlook via OAuth) through async webhooks with retries.
Team & CRM integrations Built notification and ticketing integrations with Slack, Microsoft Teams and HubSpot CRM.
Secure training pipeline RSA-signed file uploads and malware scanning (AWS EventBridge + GuardDuty) before files reach the knowledge base.

MERN Stack Developer

Ayykori Digital

Jul 2023 – Dec 20236 mos

A fintech startup developing affiliate marketplace and custom softwares for niche industries.

Affiliate management platform Built core modules for a fintech startup: product listings, campaign tracking and admin analytics, across the Node.js backend and React frontend.
Internal HRM Built leave management and attendance modules end to end, from REST APIs to the UI.
Shared foundations Reusable React and Material UI components and secure REST APIs used across both products.