Md. Farhad Billah

Md. Farhad Billah profile photo

farhadbillah2020@gmail.com

I am working as an AI Engineer at SM Technology, Dhaka, Bangladesh. AI Engineer with 2+ years of experience in OCR, voice assistants, text-to-video, and AI-powered content generation. Skilled in LLMs, RAG, FastAPI, and computer vision, with experience developing educational AI content for children. Passionate about turning research into scalable real-world solutions. Author of 3 IEEE/journal and conference papers. Japanese proficiency: JLPT N5 (日本語).

CV / GitHub / LinkedIn / YouTube / Google Scholar / Website

Education

BSc in CSE (Major: Data Science) — East West University, Dhaka, Bangladesh 2020 – 2024

CGPA 3.08 / 4.00 (≈ 85%)

Work Experience

AI Engineer — SM Technology, Dhaka, Bangladesh Dec 2024 – Present

Developed and deployed multiple AI-driven applications, including a paragraph evaluator, image processing algorithms, text-to-video generator, fitness recommendation system, and voice-powered assistant. Additionally, built intelligent solutions for content summarization, goal tracking, and personalized self-improvement.

Research Assistant — East West University, Dhaka, Bangladesh Jan 2026 – Present

Worked on a capstone project in a team of four under the supervision of Dr. Md. Sawkat Ali, Associate Professor. Led end-to-end project planning, task allocation, and timeline management to ensure timely delivery. Collaborated with stakeholders to gather requirements, manage expectations, and present progress updates.

Data Analyst Intern — Designers Do Futuro, Dhaka, Bangladesh Aug 2023 – Jan 2024

Worked on diverse Machine Learning projects involving data cleaning, transformation, preprocessing, enhancement, and visualization. Built optimized pipelines to prepare high-quality datasets for reliable model training and analysis.

Technical Skills

Programming Languages Python, C++, C, SQL
AI / ML & Generative AI ML, DL, NLP, CV, RAG, MCP, PyTorch, TensorFlow/Keras, Scikit-learn, LLMs
Frameworks & Tools FastAPI, Docker, Postman, Git, Hugging Face, OpenAI, Groq, Pydantic, Termius
Specialties (Applied & Data) OCR, Voice-to-Text, Text-to-Video, Image Processing, Recommendation Systems, Pandas, NumPy, Data Cleaning, Data Visualization, Data Mining, EDA, Statistical Modeling
Web Technologies React, HTML, CSS

Projects at SM Technology as AI Engineer

  1. Hook – AI Rizz Wingman: An AI assistant that accepts images from social media (Messenger, Instagram, LinkedIn, and dating apps) and generates intelligent, context-aware replies to help users respond appropriately and effortlessly. [App Store]
  2. MANIFEX: A paragraph evaluator that checks sentence structure and word usage, assigning marks based on accuracy. The user provides a paragraph topic, and the AI suggests required keywords that must be included. [Google Play] [Website]
  3. DFTJUNKIE: Performs image enhancement, background removal, and automatic trimming using advanced image processing algorithms. [Website]
  4. PR3DETOR FITNESS: An AI system that recommends gym exercises with YouTube tutorial videos and provides a private chat system for premium subscribers. [App Store]
  5. GREETELY: Based on the user's selected department, categories, tone of values, and employee recognition criteria, the system generates a concise email for the specified recipient and automatically sends it.
  6. AKKORD: An OCR system that extracts key contract details such as risk, summary, ownership, payment issues, termination conditions, financial terms, missing protections, and unusual clauses. Includes a chat feature for related questions.
  7. DONNA LIFE: An AI voice assistant that helps manage tasks hands-free using natural voice commands. Uses FastAPI and OpenAI models to create tasks, answer schedule questions, and offer smart, contextual assistance.
  8. Youimah: Uses Google VEO3 for text-to-video generation. The system can add background music, and characters are able to talk.

Projects at Designers Do Futuro as Data Analyst

  1. TikTok Data Visualization: A comprehensive visualization of TikTok data with proper data cleaning and preprocessing. Focuses on transforming raw data into clear, insightful, and meaningful visual analytics.
  2. Predicting User Churn in Waze: A machine learning–based approach for enhancing user retention strategies through data analysis, data preprocessing, and data visualization.
  3. FairFare: NYC TLC's Advanced Fare Estimation App by Automatidata. Trains multiple machine learning models, evaluates their performance, and selects the best-performing model for accurate fare prediction.
  4. Weather Impact on NYC Subway: Rainfall classification using multiple machine learning models to analyze patterns and improve prediction accuracy.

Research Papers

  1. Deep Learning Models for Classification of Red Blood Cells in Microscopy Images for Anemia Diagnosis : COMPAS Conference, 2024. (Digital image processing — East West University) [Link]
  2. Revisiting ONE Simulator in IoV Research: Seeing the Forest Through the Trees : IEEE Access Journal, 2025. (Capstone Projects A, B & C — East West University) [Link]
  3. ShrimpXNet: A Transfer Learning Framework for Shrimp Disease Classification with Augmented Regularization, Adversarial Training, and Explainable AI : 3ICT Conference, 2025. [Link]