I am passionate about continuously improving my knowledge and skills, and always eager to learn new technologies.

I help businesses build practical AI/ML systems, evaluate model performance, and create reliable Python-based data pipelines. My work focuses on helping clients move from messy data, manual processes, or early AI ideas to clean, measurable, and production-ready systems. Whether you need a machine learning model, data pipeline, forecasting workflow, model evaluation process, or Python-based automation, I can help you design and build a solution that is clear, reliable, and useful. I can help with: * Machine learning model development and evaluation * AI/LLM workflow testing and response evaluation * Data cleaning, transformation, and analysis * Python-based automation and backend workflows * Data pipelines for structured and unstructured data * Forecasting and prediction models * Classification, clustering, and anomaly detection * Dashboard-ready datasets and reporting workflows * API integration and model output processing I focus on practical implementation, not just technical experiments. My goal is to help you build systems that are accurate, easy to understand, maintainable, and useful for real decision-making. I work with Python, SQL, R, machine learning libraries, data visualization tools, and cloud-based data workflows. I can also support projects involving model monitoring, prompt testing, evaluation metrics, data validation, and process automation. If you are looking for someone who can connect AI/ML, data engineering, model evaluation, and Python development into one practical solution, I would be happy to help.

Need a Salesforce expert who can combine the power of AI with a scalable CRM? You're in the right place. I'm a Salesforce AI Developer with expertise in building intelligent, automated, and user-friendly Salesforce solutions that help businesses save time, improve customer experiences, and grow faster. Whether you need to customize Salesforce, implement Agentforce, integrate Einstein AI, or connect your CRM with modern AI models like OpenAI, Claude, or Llama, I can help you build a solution that delivers real business value. What I can help you with: ✅ Agentforce implementation and custom AI agents ✅ Einstein AI & Prompt Builder ✅ Salesforce development with Apex & Lightning Web Components (LWC) ✅ Flow Builder and CRM automation ✅ Sales Cloud, Service Cloud & Experience Cloud ✅ Data Cloud implementation and customer insights ✅ AI chatbots and conversational experiences ✅ OpenAI, Claude & Llama integrations ✅ LangChain & RAG-powered AI assistants ✅ Python, FastAPI, REST APIs & MuleSoft integrations ✅ Custom Salesforce apps and third-party integrations I don't just build features—I build solutions that simplify operations, automate repetitive tasks, and help your team work more efficiently. My goal is to deliver clean, scalable, and maintainable Salesforce applications that continue to add value as your business grows. I believe clear communication, reliable delivery, and understanding your business goals are just as important as writing great code. You'll always know where your project stands, and I'll work closely with you to ensure the final solution meets your expectations. If you're looking for a dependable Salesforce AI developer who understands both enterprise CRM and the latest AI technologies, I'd love to hear about your project. Let's build something amazing together. Core Skills: Salesforce • Agentforce • Einstein AI • Data Cloud • Apex • Lightning Web Components (LWC) • Flow Builder • Sales Cloud • Service Cloud • Experience Cloud • OpenAI • Claude • Llama • LangChain • RAG • AI Agents • AI Chatbots • CRM Automation • Python • FastAPI • REST API • SOAP API • MuleSoft

I build production-ready LLM systems and agentic AI pipelines that automate complex workflows. Top Rated on Upwork | 20+ projects in multi-agent AI, generative models & data engineering with PyTorch. Clients come to me when they need AI systems that actually ship — not prototypes. Whether it's a multi-agent orchestration layer, a custom RAG pipeline, a diffusion model for audio/image generation, or a data engineering workflow at scale, I handle the full cycle from architecture to deployment. What I work on: → Agentic AI & Multi-Agent Systems (LangChain, LangGraph, Groq, tool-use agents) → LLM Integration & RAG (retrieval pipelines, fine-tuning, prompt engineering) → Generative AI — Audio & Image (diffusion models, voice synthesis, latent space modeling) → ML Model Development (PyTorch, TensorFlow, classification, NLP, CV) → Data Engineering & Pipelines (PySpark, AWS SageMaker, ETL, 100GB+ datasets) Recent work includes a 7-layer multi-agent data processing engine, a latent diffusion model for synthetic medical audio generation (reduced FAD from 18.82 → 14.71), and a natural language chatbot interface for Grafana dashboards. I communicate clearly, deliver on time, and document everything. I don't disappear mid-project and I don't overpromise scope. If something's not feasible, I'll tell you upfront. If you're building something serious in AI — let's talk. Send me a message and I'll usually reply within a few hours.

I build production-ready Computer Vision, Generative AI, and Data Science systems that solve real business problems, from object detection pipelines and LLM-powered RAG applications to intelligent AI agents that automate complex workflows. Top Rated AI/ML Engineer with 5+ years delivering end-to-end solutions using PyTorch, YOLO, LangChain, OpenAI API, and Anthropic API. I've shipped defect detection models for manufacturing floors, built RAG chatbots that sit on top of company knowledge bases, designed fraud detection pipelines for fintech clients, and deployed AI agents that replaced hours of manual work. Every project I take on goes from problem to working, deployed system, not just a Jupyter notebook. I communicate clearly, deliver on time, and care about your outcome as much as the tech behind it. 👁️ Computer Vision: • Object Detection & Tracking (YOLOv8, YOLOv9, YOLOv11, SSD, Faster R-CNN, DETR) • Image Segmentation (Semantic, Instance, Panoptic, SAM, SAM 2) • OCR & Intelligent Document Processing (TesseractOCR, EasyOCR, PaddleOCR, DocTR) • Face Recognition, Facial Attribute Analysis & Emotion Detection • Vision Language Models (CLIP, BLIP, Florence, GPT-4V, Gemini Vision) • Image Enhancement, Super-Resolution & Visual Search • Video Analysis & Real-time Object Tracking (DeepSORT, ByteTrack, BoT-SORT) • AI Image Generation & Editing (Stable Diffusion, ControlNet, Inpainting) • CNNs, Vision Transformers (ViT), EfficientNet & Transfer Learning • Visual Inspection & Defect Detection for Manufacturing & Quality Control • Medical Image Analysis & Document Parsing • Multimodal AI Systems (combining vision, text & audio) 🤖 Generative AI & LLMs: • RAG (Retrieval Augmented Generation) Pipelines & Knowledge Bases • LLM Fine-tuning (LoRA, QLoRA, PEFT, RLHF, DPO) • Prompt Engineering & Optimization • OpenAI API (GPT-4, GPT-4o, GPT-5), Anthropic API (Claude), Google Gemini • Open-source LLMs (Llama 3, Mistral, Qwen, Phi, DeepSeek, Gemma) via Hugging Face Transformers • Custom AI Chatbot Development & Virtual Assistants • LangChain, LlamaIndex, Haystack • AI Integration into Existing Systems & Workflows • Speech-to-Text (OpenAI Whisper) & Text-to-Speech (TTS, ElevenLabs) • Embedding Models & Vector Search • Document AI: Extraction, Parsing & Intelligent Processing 📊 Data Science & Machine Learning: • Exploratory Data Analysis (EDA) & Feature Engineering • Predictive Modeling, Forecasting & Predictive Analytics • Classical ML (XGBoost, LightGBM, CatBoost, Random Forest, SVM, KNN) • Time Series Analysis, Forecasting & Anomaly Detection • Classification, Regression & Clustering • Model Evaluation, Selection & Hyperparameter Tuning • Data Preprocessing, Cleaning & Wrangling • Data Visualization & Dashboards (Matplotlib, Seaborn, Plotly) • Fraud Detection & Risk Analysis • Recommendation Systems 🔄 AI Agents & Automation: • Agentic AI Workflows (LangGraph, CrewAI, AutoGen, Agno) • Multi-Agent Systems & Tool-Use Agents • Model Context Protocol (MCP) Integrations • AI-powered Workflow Automation (n8n, Make, Zapier) • End-to-end Pipeline Orchestration & AI Integration • Autonomous Task Execution & Decision-Making Systems • Voice AI Agents (VAPI, Bland AI) ✅ Tech Stack: ➼ Languages: Python ➼ ML/DL Frameworks: PyTorch, TensorFlow, Keras, Scikit-Learn, Hugging Face Transformers ➼ Computer Vision: OpenCV, Ultralytics (YOLO), PIL, TesseractOCR, EasyOCR, PaddleOCR, Roboflow ➼ GenAI / LLM: LangChain, LlamaIndex, LangGraph, OpenAI API, Anthropic API, Hugging Face ➼ Data & Analytics: Pandas, NumPy, Matplotlib, Seaborn, Plotly, Scipy ➼ API & Deployment: Flask, FastAPI, Docker, Streamlit, Gradio ➼ Cloud: AWS (SageMaker, EC2, S3, Lambda), Azure, GCP ➼ Databases: Pinecone, ChromaDB, Faiss, Weaviate, PostgreSQL, MongoDB ➼ MLOps: MLflow, Weights & Biases, Model Versioning & Monitoring ➼ IDE: VS Code, PyCharm, Jupyter Notebook, Google Colab 🚀 Why work with me: Production-Focused: I don't just prototype, I ship deployed, working systems that integrate with your workflow and deliver measurable value. Results-Driven: Every project starts with your business goal. Whether it's automating visual inspection, building a smart chatbot, detecting fraud, or extracting insights from documents, the model is a means to your outcome. Modern & Current: The AI field moves fast and so do I. I stay on top of what actually works, LLMs, RAG, AI agents, vision-language models, multimodal AI, so your solution uses cutting-edge approaches, not outdated ones. Clear Communication: Top Rated with a strong Job Success track record. I communicate clearly, meet deadlines, and treat every project as a real partnership. 🤝 Let's work together to push the boundaries of what's possible!

With 4+ years in AI, I bridge the gap between advanced research and production engineering. I don't just train models; I build the scalable infrastructure to serve them. My expertise takes custom neural networks from mathematical theory to fully deployed, production-grade APIs on Kubernetes clusters. 🛠️ Core Technical Stack 1. Generative AI & Agents Tech: LLMs, AI Agents, RAG, LiveKit, TTS, STT, LangGraph, MCP, n8n workflows. Memory & Context: mem0 (Memory Management), RAG pipeline optimization. 2. Deep Learning & Neural Networks Core: PyTorch, Stable Diffusion, Transformer, 3D U-Net, IP-Adapter, ControlNet, ArcFace. 3. Backend API Development & Real-Time Streaming Frameworks: Expert in FastAPI (Async/High-concurrency) and Flask for serving ML models. Protocols: RESTful APIs, WebSockets, gRPC, and Swagger/OpenAPI documentation. Streaming: LiveKit, WebRTC, Socket.io for low-latency interactive AI. 4. MLOps, DevOps & Big Data CI/CD & Monitoring: GitHub Actions, Sentry (Error Monitoring), Automated Testing Pipelines. Infrastructure: Docker, Kubernetes (Scalable Clusters). Data Engineering: Kafka, Debezium, PySpark for real-time data streaming. HPC: Distributed training (Multi-node/Multi-GPU) on A100 & H100 clusters. 📂 Project Showcase 🤖 Intelligent Voice Agent & RAG System Role: Led the upgrade of a Voice Agent (v0.x to v1.x), serving the RAG pipeline via optimized FastAPI endpoints. Agentic Workflow: Implemented long-term memory using mem0 and integrated n8n workflows via an MCP server. Ops: Deployed on Kubernetes with GitHub Actions for CI/CD and Sentry for real-time error tracking. 🎥 Real-Time Talking Avatar Role: Integrated LiveKit (WebRTC) with Wav2Lip for a live-streaming avatar solution. Performance: Achieved sub-200ms latency on GPU clusters using optimized socket streaming. Impact: Reduced deployment costs by 50% through rigorous model pruning, quantization, and fine-tuning. 🚀 Stable Video Diffusion & HPC Training Role: Developed distributed training code for complex architectures (3D U-Net, IP-Adapter, ControlNet) using DensePose and 3D pose datasets. Infrastructure: Trained on 4x A100 (single node) and 2x H100 (multi-node) clusters. Result: Achieved 40% faster convergence through multi-node optimization compared to single-GPU training. 🗳️ Election Prediction System (Oman Shura Election) Role: Engineered a real-time election analysis pipeline using Debezium, Kafka, and PySpark. Modeling: Deployed Transformer-based models on 4x A10 GPUs to forecast polling percentages for 63 regions. Visualization: Created interactive dashboards visualizing 6 years of historical voting data. 💬 Speech-to-Text (Whisper ASR) Role: Trained and deployed OpenAI’s Whisper model using the Libri-speech dataset. Impact: Created a high-availability API pipeline capable of handling diverse audio inputs.

🚀 AI-Driven Energy Forecasting | Data Science | Machine Learning Are you looking to optimize energy operations with AI, or need accurate demand forecasting? I’m a Data Scientist with a Master’s degree and over 4 years of experience in applying Machine Learning, Time Series, and Predictive Modeling to energy systems and data-intensive problems. ✅ My expertise includes: 📊 Time Series Forecasting – energy demand, pricing, weather, streamflow 🔍 Predictive Modeling – regression, classification, model tuning ⚡ AI in Energy Systems – load prediction, energy trading optimization 📈 Data Analytics & Visualization – insights that drive real-world decisions 🧠 Deep Learning & Reinforcement Learning – for autonomous systems and intelligent control 🛠️ Project Design & Consulting – roadmap creation for scalable AI solutions I’ve worked on projects involving: Electricity price and demand forecasting for utilities ML-based streamflow prediction Optimization of energy portfolios using RL Data visualization dashboards using Python (Plotly, Dash, etc.) 🔧 Tools/Tech: Python, Scikit-learn, TensorFlow, Keras, pandas, NumPy, XGBoost, Prophet, SQL, Power BI Let’s discuss how I can bring real intelligence to your energy or analytics challenges.

My name is Marouane MARHOUN, and I am a Machine Learning and Deep Learning enthusiast based in Morocco. I specialize in providing efficient solutions utilizing machine learning and deep learning models, whether implemented from scratch, based on academic papers, or through fine-tuning existing models. I possess extensive experience in developing, training, evaluating, and deploying deep learning models across various domains, including Natural Language Processing (NLP) and computer vision. My project portfolio encompasses a range of tasks such as clustering, image recognition, CSV text classification, time-series forecasting, classification, segmentation, and detection. I hold a Master's Degree in Signal Processing and Machine Learning, complemented by multiple certifications from esteemed platforms such as Coursera, Kaggle, IBM, and CognitiveClass. My areas of expertise include: ✔️ Machine Learning ✔️ Deep Learning ✔️ Exploratory Data Analysis ✔️ Digital Image Processing ✔️ Optimization ✔️ Data Mining ✔️ Data Visualization ✔️ Web Scraping For further insights into my work, please visit my GitHub repository: github.com/MarMarhoun. I welcome the opportunity to discuss your project in detail and am eager to assist you with any challenges you may face.

Struggling with building powerful AI systems that actually deliver business results? I specialize in turning complex Machine Learning (ML), Deep Learning (DL), and Large Language Models (LLMs) into simple, scalable, business-focused solutions. As a Top Rated AI Developer with 40+ successful projects, I help businesses: Predict outcomes and optimize operations with custom ML models Deploy real-world AI solutions — not just theoretical prototypes Build AI-powered chatbots, automation systems, and smart assistants Save time, increase revenue, and stay ahead of competitors through intelligent automation What I Offer: ✅ End-to-End Machine Learning Solutions (preprocessing to deployment) ✅ Advanced Deep Learning Systems (CNNs, GANs, Transformers) ✅ NLP Solutions (chatbots, text analysis, language modeling) ✅ LLM Integration (GPT-4, Claude, Llama2, Gemini, Falcon 7B) ✅ Automation & Workflow Orchestration (Make.com, N8N, LangChain, CrewAI) ✅ AI-Powered Computer Vision (YOLOv8, SAM, EfficientNet, Unet, MediaPipe) ✅ Real-Time ML Deployments (Flask, Django, AWS, GCP) ✅ IoT & Edge AI Solutions (Jetson Nano, Raspberry Pi, Arduino) ✅ Scalable AI Infrastructure on Cloud Why Work With Me? Clear and Responsive Communication Cutting-edge AI Techniques for Maximum Accuracy Timely Project Delivery & Full Testing/Validation Business-Centric Solutions, Not Just Code Tech Stack: Python | TensorFlow | PyTorch | Hugging Face | OpenCV | Flask | Django | AWS | Make.com | N8N | PostgreSQL | Pinecone | Faiss | Chroma | Qdrant Ready to transform your data into business intelligence or automate your workflows? Let’s connect and turn your AI vision into reality. Message me today!

I bring over 2 years of hands-on experience at the Singapore University of Technology and Design, where I’ve built real-world intelligent systems using Deep Learning, Generative AI, and Reinforcement Learning — and I’m just getting started. I’m a Computer Science graduate, and I’ve spent over 5 years personally learning, exploring, and applying AI across different fields. From tracking athletes and guiding robots to reading license plates and monitoring construction site progress, I specialize in turning creative ideas into practical, and high-impact ML solutions. Accomplishments: ⭐️ IBM Hackathon Winner – Built a real-time deep learning system for cricket pose estimation. ⭐️ ANPR | FRS for KAVACH Hackathon (Top 5) – Developed YOLOv8 based multilingual license plate + OCR and passenger detection solution. ⭐️ 360° Construction site Monitoring – Deployed a panoramic image segmentation system for tracking and analyzing construction site progress. ⭐️ Generative RL for Robotics – Created multi-agent systems that enable helps robots figure out how to assemble things @SUTD. ⭐️ NLP for EdTech – Built transformer-based machine translation solution and Chatbot for a global learning platform. Core Skills & Tools: ✔Machine Learning with Pytorch, Tensorflow, Scikit Learn, Unity 3D (ML - Agents), Unreal Engine 5.0 ✔NLP with Transformer models, RNN, Attention models and Text data preprocessing with Spacy. ✔Image Segmentation with U-Net, PIX2PIX GAN, Spherical U-Net (360 degree image segmentation), 3D Space Segmentation (with 3DGS, NeRF, GARField etc). ✔Object Detection and Image Classification with YOLO, Efficient Nets, OneShot Learning etc. ✔Generative Modeling with RL (Deep Q Learning, PPO, DDPG, Multi-Agent RL, Unity ML-Agents) ✔Additional Skill - Animation and Designing with blender and UE My research expertise ensures timely delivery and high-quality results. I believe in strong communication, idea-sharing, and building solutions that actually make a difference. Let’s bring your AI vision to life.