Michele isn't taking new orders for this project right now. Here are some similar projects to explore.
You will get a fine-tuned computer-vision classification model ready to be deployed
Project details
šš«š¢š§š š²šØš®š« ššššš¬šš šš§š š šš š šš®š¬ššØš¦, š”š¢š š”š„š²-šØš©šš¢š¦š¢š³šš, šš„šš¬š¬š¢šš¢šššš¢šØš§ š¦šØššš„ ššØš« š²šØš®š« ššØš¦š©š®ššš«-šÆš¢š¬š¢šØš§ šš©š© š¢š§ šš¬ š„š¢ššš„š šš¬ š ššš²š¬.
With our proprietary system you get a model trained by the latest research in deep-learning architectures and machine-learning training procedures. The models we provide are ready for deployment, and come with code and documentation to easily run predictions on a mobile device or a cloud service.
What you'll get:
ā¦ A state-of-the-art classification model that can be imported in 3 different frameworks: PyTorch, TensorFlow and ONNX
ā¦ Example code and documentation explaining how to load and run inference on the model in each of the aforementioned frameworks
ā¦ A detailed report obtained on an out-of-sample test set that shows how the model is performing
Optionally, you can have:
ā¦ A highly-optimized model to run blazing-fast predictions in the cloud (or on processors without a floating-point unit) and up to 40x smaller than the size of a standard neural network. (Available with the "Model Optimization" add-on.)
With our proprietary system you get a model trained by the latest research in deep-learning architectures and machine-learning training procedures. The models we provide are ready for deployment, and come with code and documentation to easily run predictions on a mobile device or a cloud service.
What you'll get:
ā¦ A state-of-the-art classification model that can be imported in 3 different frameworks: PyTorch, TensorFlow and ONNX
ā¦ Example code and documentation explaining how to load and run inference on the model in each of the aforementioned frameworks
ā¦ A detailed report obtained on an out-of-sample test set that shows how the model is performing
Optionally, you can have:
ā¦ A highly-optimized model to run blazing-fast predictions in the cloud (or on processors without a floating-point unit) and up to 40x smaller than the size of a standard neural network. (Available with the "Model Optimization" add-on.)
What's included $1,400
These options are included with the project scope.
$1,400
- Delivery Time 7 days
- Number of Revisions 1
- Number of Model Variations 3
- Model Validation/Testing
- Model Documentation
- Data Source Connectivity
- Source Code
Optional add-ons
You can add these on the next page.
Fast 3 Days Delivery
+$700
Additional Revision
+$400
Model Optimization
(+ 1 Day)
+$200
Data Cleaning
(+ 1 Day)
+$400
Data Formatting
+$100Frequently asked questions
4 reviews
(4)
(0)
(0)
(0)
(0)
This project doesn't have any reviews.
CH
Carl H.
May 15, 2023
Document Classification and Extraction
Michele provides great high quality work! Forward thinker.
SB
Sam B.
Oct 5, 2022
Machine Learning Engineer for ML Marketplace Prototype
SP
Stephane P.
Jan 18, 2021
Machine Learning / Vision AI Engineer in an advisory role
Very knowledgeable in the ML/AI field and provided some very valuable feedback for our project.
SE
Sam E.
Dec 21, 2015
Audio Storage, Recognition, and Machine Learning
About Michele
Senior ML Engineer | Caltech | Computer Vision & Quant Systems
100%
Job Success
Playa del Carmen, MexicoĀ - 11:04 am local time
My expertise is evenly split between engineering complex visual intelligence (object detection, sensor fusion, edge AI) and designing predictive quantitative pipelines for derivatives pricing and alpha generation.
Here is a non-exhaustive list of my core competencies and technical skills:
āŗ MACHINE LEARNING & DEEP LEARNING (14+ Years)
ā¢ ARCHITECTURE DESIGN: Transformer Models (LLMs, Fine-Tuning, Vision Transformers), State Space Models (Mamba, SSMs), Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), Transformer Models (LLM and VT), Autoencoders, and Graph Neural Networks (GNN).
ā¢ TECHNIQUES: Supervised & Unsupervised Learning, Reinforcement Learning, Transfer Learning, Hyperparameter Tuning, Anomaly Detection, Feature Engineering, Feature Selection, and AI Bias Mitigation.
ā¢ CLASSICAL ML & DATA ANALYSIS: Gradient Boosting (XGBoost, LightGBM, AdaBoost), Random Forests, Support Vector Machines (SVM), Principal Component Analysis (PCA), and non-linear data analysis and dimensionality reduction (t-SNE, UMAP) using Scikit-Learn.
ā¢ FRAMEWORKS & TOOLING: PyTorch, TensorFlow, Keras, NumPy, pandas, and Databricks MLflow for model tracking and lifecycle management.
ā¢ STATISTICAL MODELING: Bayesian Analysis & Statistics (PyMC), Gaussian Processes, and robust time-series forecasting.
āŗ QUANTITATIVE FINANCE & ALGORITHMIC TRADING
ā¢ TIME-SERIES ANALYSIS: Extracting predictive signals from noisy, granular datasets (up to 1-second resolution via Databento, dxLink) and leveraging zero-shot Time-Series Foundation Models (Amazon Chronos, TimeGPT).
ā¢ DERIVATIVES & VOLATILITY: Core expertise in equity options pricing, modeling volatility surfaces (Dupire, Heston, VIX term structure), and managing higher-order Greeks (Vanna, Charm, Volga).
ā¢ MARKET MICROSTRUCTURE & RISK: Analyzing Limit Order Book (LOB) dynamics and engineering custom, leak-proof backtesting engines utilizing robust time-series cross-validation methodologies to strictly prevent data leakage.
āŗ COMPUTER VISION (14+ Years) & NLP (10+ Years)
ā¢ COMPUTER VISION: Expertise spanning Image Classification, Object Detection & Tracking (e.g., YOLO), Video Tracking, Semantic Segmentation, Face Detection, and Sensor Fusion.
ā¢ NLP & GENERATIVE AI: Text Classification, AI Image Generation (Stable Diffusion), and API integration with Large Language Models (Claude, Gemini, ChatGPT).
ā¢ MODEL OPTIMIZATION & DEPLOYMENT: End-to-end model testing, optimization, and deployment on edge devices using OpenCV, TensorFlow Lite, NVIDIA Jetson, and Raspberry Pi.
āŗ CLOUD INFRASTRUCTURE, BIG DATA, & MLOps
ā¢ CLOUD PLATFORMS: Advanced deployment utilizing AWS (Amazon EC2, Amazon ECS, Amazon S3, Amazon SageMaker, Amazon API Gateway) and Google Cloud Platform (including Vertex AI).
ā¢ BIG DATA & ANALYTICS: Utilizing Hadoop, Apache Spark MLlib, and Databricks for large-scale cluster computing and data analysis.
ā¢ SYSTEMS ENGINEERING: Python (primary research stack) and C++ / Java (low-latency execution), alongside Bash scripting and SQL / NoSQL Database management (e.g., high-performance caching with Redis).
ā¢ DEVOPS: Containerization with Docker, continuous deployment, version control with Git, and enforcing rigorous Data Science Consultation practices.
Are you looking to build a highly optimized algorithmic trading engine or deploy state-of-the-art AI infrastructure? Send me a message outlining your project, and we can discuss the optimal architecture and execution strategy.
Steps for completing your project
After purchasing the project, send requirements so Michele can start the project.
Delivery time starts when Michele receives requirements from you.
Michele works on your project following the steps below.
Revisions may occur after the delivery date.
Data Analysis
After you send us the link to your dataset as specified in the requirements, we'll analyze the data and get back to you within 12 hours.
Data Cleaning and Formatting
If you chose to add the "Data Cleaning" or "Data Formatting" add-on, we structure the dataset in a format that is compatible with our system, and remove all the samples that would introduce noise into the model.