We'll get to know your use case and requirements in detail. We'll find out if a quick pre-trained open-source model is a fit for your requirements or we may need to fine-tune or train from scratch a model.
You will get a RESTful API of your trained and deployed machine learning model.
- Delivery Time 18 days
- Number of Revisions 2
Fund the project upfront. Hamza gets paid once you are satisfied with the work.
Project detailsBased on your requirements, I'll create:
A cost effective RESTful API of your prediction model (machine learning model) which you can integrate into your app.
✅ "Starter Tier": If you've a machine learning model, I'll deploy that quickly.
✅ If you just know your use case yet do not have (or know of) any machine learning model fit for your project. I'll find (or train) the model for you. Specifically, I'll do the following before deploying:
✔️ "Standard Tier": Use a pre-trained model based on your use case (e.g. transformer model such as BERT, GPT) OR
✔️ "Advanced Tier": Fine-tune the pre-trained model on your custom dataset OR
✔️ "Advanced Tier": Train a model from scratch.
✅ If you need a use-interface (not just the API) for your app, I'll write the UI in React with API integrated.
- 1Gather Requirements
- 2Test Model locally
Before deploying anything, we'll test the model locally. Perhaps, using Google Colab. After this step, we'll be certain that the model meets our requirements (i.e. given this input, returns this output), and is ready to be deployed.
- 3Deploy the Model in Production on AWS
We'll determine the right infrastructure for deployment. ✅ Solution 1: Deploy on Amazon Lambda ✅ Solution 2: Deploy on Amazon Sage Maker. Each solution has its pros and cons and depends on the requirements.
- 4Test the deployed Model
Finally, we'll test the deployed model and make any changes if necessary.
- 1Which best describes your use-case (requirement)?
- 2What are your requirements?
Frequently asked questions
I am the author of the book titled "Practical Machine Learning" which is available publicly. The link and the description of the book is given further below in "Other Experiences".
I believe if one has the right programming and problem-solving ability then no problem is unsolvable. I love solving challenging problems.
Deploying my strong algorithmic, software engineering and mathematical background, I will enthusiastically work on your Machine Learning projects.
👍 “He was very cooperative and also have good communication skills. He was also very helpful with bringing his own ideas and experience to the table. He's a valuable asset to any team.”
👍 “He is an extremely talented AI expert. Never miss an opportunity to work with him.”
👍 “Hamza is a top notch data scientist with good insight into the best path to create machine learning capabilities. He is good at communications, is diligent and is capable of delivering within timelines.”
✅ Machine Learning Engineer
✔️ Training prediction models
✔️ Natural Language Processing, Understanding and Generation (NLP/NLU/NLG)
✔️ Computer Vision (CV)
✔️ Transformer Models
✔️ Auto-encoding Models (BERT, RoBERTa and others)
✔️ Auto-regressive Models (GPT, GPT-2 and others)
✅ DevOps/MLOps Engineer
✔️ Deploying and monitoring production systems. Handling infrastructure.
✔️ Creating cost-effective RESTful APIs of your Machine Learning prediction models.
⚙️ AWS Lambda
⚙️ Amazon Sage Maker
⚙️ Google Cloud platform
⚙️ TensorFlow Serving
✔️ As a DevOps engineer, I’ll deploy your machine learning models in a cost efficient manner that are scalable and maintainable. Contrary to what GCP or AWS tells you, you don’t always need expensive GPUs for prediction systems (i.e. at inference time). I’ll give you a RESTful API(s) which you can integrate into your app.
✔️ If you need more than an API, I’ll write the User-Interface for you as well.
⚙️ React JS
✅ Data Engineer
✔️ Building data pipelines. Extracting features for machine learning models.
✔️ Technical Writing/Presenting.
✔️ Label Studio
✔️ Label Studio with ML backend