AI Engineer  |  Generative AI Specialist  |  LLM Systems Architect  |  AWS Cloud AI

Computer Science graduate from American International University-Bangladesh, specializing in Data Science and Machine Learning.

I design, build, and deploy production-grade AI systems focused on:

• Generative AI Applications
• Large Language Model (LLM) Engineering
• Retrieval-Augmented Generation (RAG)
• Agentic AI & Autonomous Workflows
• Cloud-Native AI Deployment (AWS)
• Secure & Reliable AI Systems

My engineering approach emphasizes scalability, reliability, structured orchestration, security, and measurable business impact.

Oct 2025 – Feb 2026

• Engineered a production-ready LLM-powered conversational commerce system integrated with APIs for end-to-end food ordering.
• Reduced customer support workload by 60% through AI-driven automation.
• Designed and implemented agentic workflows using LangChain and OpenAI APIs.
• Conducted prompt injection testing and mitigation to improve AI system security.
• Deployed scalable AI services using AWS Lambda, EC2, and DynamoDB.
• Implemented multi-platform AI deployment (Messenger, WhatsApp, LinkedIn, Web) via webhook architecture.
• Structured AI knowledge bases and optimized response performance for SaaS deployment.

Amazon Bedrock   AWS Lambda   AWS EC2   AWS DynamoDB   Cloud Deployment & Scalability Optimization

Tech Stack: LangChain · LangGraph · OpenAI API · Python · Pandas

• Architected a structured LLM-powered transactional workflow system.
• Implemented strict execution pipeline: Menu → Cart → Confirmation → Checkout → Data Persistence.
• Built custom tool-based reasoning modules for menu validation, order verification, and bill calculation.
• Integrated GPT function-calling to minimize hallucinations and enforce structured outputs.
• Designed session management, edge-case handling, and persistent order logging for production reliability.

Tech Stack: R · NLP · LDA · arXiv API

• Automated ingestion of 10,000+ research abstracts via arXiv API.
• Engineered NLP preprocessing pipeline (lemmatization, tokenization, stop-word removal).
• Implemented Latent Dirichlet Allocation (LDA) using Gibbs Sampling.
• Delivered structured topic probability distributions for analytical insights.

Enterprise-Grade LLM Systems        Scalable Agentic AI Architectures
Secure AI Deployment                RAG Optimization
Cloud-Native AI Infrastructure

Open to AI Engineering roles, research collaborations, and impactful projects.

almaruf0006@gmail.com  |  LinkedIn  |  GitHub