Skip to content
View Rhythm05Roy's full-sized avatar
4 followers · 5 following
  • Daffodil International University
  • Bangladesh
  • LinkedIn in/rhythm-roy

Achievements

Achievement: Pull Shark

Achievements

Achievement: Pull Shark

Highlights

  • Pro

Block or report Rhythm05Roy

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Rhythm05Roy/README.md
Typing SVG

🔭  Currently building production AI systems at Betopia Group  ·  🎓  Seeking PhD opportunities in the US (AI/ML)  ·  📍  Dhaka, Bangladesh


🛠️ Tech Stack

Languages

Python C++ Java SQL Bash JavaScript

ML / DL Frameworks

PyTorch TensorFlow Keras HuggingFace Scikit-learn

LLMs & Agentic AI

LangChain OpenAI FastAPI

MLOps & Cloud

Docker Kubernetes AWS GCP Azure MLflow Git Linux

🚀 What I Build

🤖 LLM & Agentic Systems

  • Production RAG pipelines with hybrid retrieval (BM25 + semantic)
  • Multi-agent orchestration with LangGraph & CrewAI
  • Real-time voice AI agents with tool-calling (Twilio + ElevenLabs + Whisper)
  • Parameter-efficient fine-tuning with LoRA / QLoRA on low VRAM

⚡ Efficient Deep Learning

  • Model compression — distillation, pruning, quantization (INT8/4-bit)
  • Adversarial robustness & test-time adaptive purification
  • Visual token pruning for vision-language models
  • Curriculum reinforcement learning for small language models

🏗️ MLOps & Production AI

  • End-to-end ML pipelines: ingest → validate → train → deploy
  • Containerized workflows with Docker & Kubernetes
  • Cloud artifact management — AWS S3, MongoDB, MLflow, DVC
  • GAN architectures deployed to production

👁️ Computer Vision & Multimodal AI

  • Spatiotemporal video analytics with YOLOv8 + optical flow
  • Radiology image analysis → structured medical report generation
  • Classical CV pipelines: edge detection, homography, OCR
  • Multi-modal conditioned image generation orchestration

📊 GitHub Stats


📈 Contribution Activity

Let's Connect

Email LinkedIn Portfolio


Profile Views


Pinned Loading

  1. Sina_laut Sina_laut Public

    An asynchronous FastAPI backend framework for constraint-conditioned vision synthesis, leveraging background workers and structured metadata validation to orchestrate deterministic multimodal image…

    Python

  2. speaklar_rag speaklar_rag Public

    A high-performance, asynchronous Multimodal Retrieval-Augmented Generation (RAG) system featuring hybrid search (dense + sparse), coreference resolution, and multi-turn vision-language question ans…

    Python

  3. Vehicle-Insurance-MLOps-project Vehicle-Insurance-MLOps-project Public

    A production-grade, containerized MLOps infrastructure featuring automated data ingestion, validation schemas, transform pipelines, and tracking configurations integrated with AWS S3 and MongoDB.

    Python

  4. Video-Feature-Extractor Video-Feature-Extractor Public

    An extensible spatiotemporal video profiling pipeline that integrates optical flow motion magnitude, pixel-differential shot detection, YOLOv8 spatial object mapping, and Tesseract OCR text extract…

    Python

  5. LLama_fine_Tune LLama_fine_Tune Public

    A memory-efficient Parameter-Efficient Fine-Tuning (PEFT/QLoRA) framework using Unsloth to adapt autoregressive Large Language Models (LLaMA 3.1) for low-resource empathetic dialogue generation.

    Jupyter Notebook

  6. Voice-Receptionist-Agent Voice-Receptionist-Agent Public

    A stateless FastAPI microservice orchestrating low-latency, real-time conversational voice agents for automated home services, featuring TwiML webhooks, ElevenLabs Arabic TTS, multi-LLM routing, an…

    Python