AI-Powered WhatsApp Automation
A WhatsApp AI agent built with n8n and Gemini that automates responses to user queries, with conversation memory and external tools for contextual replies and information retrieval.
I'm a Computer Science and Engineering graduate from BRAC University with hands-on experience in data analysis, web application development and workflow automation. I enjoy taking messy, scattered data and turning it into clear answers that teams can act on.
Currently I work as a Junior Data Analyst at CarryBee, where I build ETL and reporting workflows with Python, PostgreSQL, n8n and Google Apps Script, analyze merchant churn and delivery performance, and develop KPI dashboards in Metabase and Power BI.
My interests lie in AI-powered automation, business analysis and data engineering: understanding how a business runs, measuring it with the right KPIs, and building the pipelines, AI agents, dashboards and integrated applications that streamline operations and support data-driven decisions.
SQL, Power BI, Metabase, Looker Studio, Excel, Google Sheets
Python, JavaScript, PostgreSQL, MySQL, MongoDB
Pandas, NumPy, Matplotlib, Scikit-learn, SQLAlchemy
n8n, Google Apps Script, Sheets API, Python ETL pipelines
FastAPI, React, Node.js, Git, GitHub
A WhatsApp AI agent built with n8n and Gemini that automates responses to user queries, with conversation memory and external tools for contextual replies and information retrieval.
A web dashboard comparing merchant success rates, return rates, revenue and parcel aging. Includes interactive filters, forward and reverse aging views, CSV exports and data validation alerts.
Analyzed city-level delivery outcomes and rider efficiency to identify operational performance gaps, then built SQL-driven Power BI reports to monitor delivery KPIs and SLA breach patterns.
A classification pipeline covering data preprocessing, encoding, model training and evaluation. Compared multiple classifiers and identified Random Forest as the best-performing model.
A Bangla visual question answering system combining image, speech and text inputs, exploring vision-language models to support crop disease diagnosis.
A full-stack food ordering app with authentication, menus, ratings and order tracking, plus backend workflows for order handling and platform interactions.
Open to Data Analyst, Business Analyst and AI & Automation Engineer roles, as well as collaborations and freelance work.