Senior ML Platform Engineer and AI Architect

Josh Hamet Staff AI/ML Engineer

****.*******@*****.*** 313-***-**** Saline, Michigan

https://www.linkedin.com/in/josh-hamet/

Professional Summary

Senior Machine Learning Engineer with strong expertise in designing and deploying scalable AI systems using Python and machine learning frameworks including PyTorch, TensorFlow, Keras, Scikit-Learn, XGBoost, and LightGBM. Experienced in developing Large Language Model (LLM), Natural Language Processing (NLP), and chatbot solutions, building intelligent applications such as semantic search, conversational AI systems, and transformer-based language models using tools like Hugging Face Transformers, spaCy, and NLTK. Proven background in building production AI platforms with FastAPI and Flask RESTful APIs, implementing MLOps pipelines, and deploying scalable machine learning services on AWS, Azure, and Google Cloud Platform using Docker and Kubernetes.

Professional Experience

Block — ML Platform Tech Lead / Staff Machine Learning Engineer July 2023 – Present

● Lead the architecture and development of scalable machine learning platforms used by engineering teams for training, evaluating, and deploying production-grade AI models.

● Design and implement distributed ML pipelines using Python and machine learning frameworks including PyTorch, TensorFlow, Keras, Scikit-Learn, XGBoost, LightGBM, and Spark MLlib.

● Develop advanced Large Language Model (LLM) solutions including conversational AI systems, enterprise chatbots, document understanding pipelines, and AI-powered developer productivity tools.

● Build scalable Natural Language Processing (NLP) pipelines using Hugging Face Transformers, spaCy, NLTK, Gensim, and OpenNLP for tasks such as entity recognition, sentiment analysis, semantic search, and language understanding.

● Develop backend services using Python frameworks including FastAPI, Flask, and Django, building highly scalable RESTful APIs to serve machine learning models and AI services.

● Deploy and manage ML services in cloud environments including AWS, Microsoft Azure, and Google Cloud Platform, leveraging services such as AWS SageMaker, Azure ML, and Google Vertex AI.

● Containerize machine learning services using Docker and orchestrate scalable deployments using Kubernetes to support high-availability AI systems.

● Implement MLOps pipelines using Kubeflow, MLflow, Apache Airflow, Jenkins, and GitHub Actions to automate training, model validation, deployment, and monitoring workflows.

● Design distributed data processing and feature engineering pipelines using Apache Spark, Hadoop, Ray, and Dask to process large-scale structured and unstructured datasets.

● Utilized SQL for querying and optimizing datasets across multiple databases to support machine learning model training and evaluation.

● Implement Retrieval-Augmented Generation (RAG) architectures using vector databases such as FAISS, Pinecone, and Weaviate to enhance LLM-powered knowledge retrieval and chatbot capabilities.

● Leveraged Snowflake for high-performance, scalable data storage and processing of structured data to support real-time analytics and model evaluations.

● Lead software engineering best practices including Python microservice development, API design, system scalability, and high-performance backend architecture.

● Mentor engineers on deep learning, ML system design, cloud architecture, and scalable AI platform development.

Twitter — Senior Machine Learning Engineer / Machine Learning Engineer II Mar 2019 – Jul 2023

● Designed and optimized a distributed machine learning experimentation platform supporting thousands of engineers using Jupyter Notebook and Python-based ML environments.

● Engineered a high-performance virtual file system for ML workflows, improving I/O operations and accelerating machine learning development cycles.

● Built scalable machine learning models using PyTorch, TensorFlow, Keras, Scikit-Learn, XGBoost, and LightGBM for prediction, ranking, and recommendation tasks.

● Developed Python-based backend systems and microservices using FastAPI and Flask, exposing machine learning models via RESTful APIs for internal applications and production services.

● Designed large-scale NLP systems leveraging transformer-based models such as BERT, RoBERTa, GPT, and T5 for text classification, semantic similarity, and content analysis.

● Implemented LLM-powered conversational AI assistants and chatbots to support internal developer tools and automate documentation search and knowledge discovery.

● Built scalable experiment tracking and monitoring systems using MLflow, Weights & Biases, and TensorBoard to manage machine learning experimentation and reproducibility.

● Designed distributed feature engineering pipelines using Apache Spark, Hadoop, Kafka, and Airflow to process massive behavioral datasets.

● Developed cloud-native machine learning infrastructure deployed on AWS, Azure, and Google Cloud Platform, enabling elastic scaling of AI workloads.

● Containerized services using Docker and orchestrated production ML workloads using Kubernetes clusters.

● Implemented CI/CD pipelines for machine learning services using Jenkins, GitHub Actions, and GitLab CI to automate deployment and model updates.

● Built internal dashboards and tools enabling engineers to visualize ML experiment results and compare model performance.

● Developed SQL-based ETL pipelines for structured data extraction, transformation, and loading into cloud databases to support machine learning models. Criteo — Senior Software Engineer / Software Engineer II–III Nov 2016 – Mar 2019

● Designed and deployed a content-based logistic regression model using TensorFlow and Scikit-Learn to predict the relevance of search term-to-product associations in advertising systems.

● Developed large-scale machine learning pipelines using Python and frameworks such as PyTorch, TensorFlow, Keras, and Spark MLlib to train predictive advertising models.

● Implemented Python-based backend services using frameworks such as Flask and Django, exposing machine learning functionality through scalable RESTful APIs.

● Built distributed data pipelines using Apache Spark and Hadoop to process billions of user interaction records for machine learning training datasets.

● Designed recommendation and ranking algorithms using gradient boosting models, collaborative filtering, and neural network architectures to improve advertising relevance.

● Implemented Natural Language Processing models using Gensim, spaCy, Word2Vec, and FastText to analyze search queries and product descriptions.

● Developed internal model validation dashboards using React and Redux, allowing engineers to analyze model results and compare experimental model versions.

● Built scalable training workflows using cloud infrastructure on AWS, Azure, and Google Cloud Platform, enabling distributed model training and data processing.

● Containerized machine learning services using Docker and deployed scalable inference systems using Kubernetes clusters.

● Developed complex SQL queries for querying large data sets and optimizing performance across distributed cloud databases.

● Participated in the Criteo Machine Learning Bootcamp, gaining advanced knowledge in machine learning algorithms, statistical modeling, and deep learning frameworks.

● Collaborated with data scientists and ML researchers to deploy high-performance predictive models into production advertising systems.

HookLogic (Acquired by Criteo) — Software Engineer July 2015 – Nov 2016

● Developed software solutions for a large-scale e-commerce advertising platform using Python and JavaScript, enabling retailers to integrate sponsored product recommendation systems.

● Built backend services using Python frameworks such as Flask and Django to support scalable RESTful APIs used by advertising and analytics platforms.

● Implemented distributed ETL pipelines using Hadoop MapReduce and Apache Spark to process large datasets of product information from multiple retail partners.

● Developed feature engineering pipelines that generated training data for machine learning models used in advertising optimization systems.

● Implemented data transformation systems to normalize product metadata from multiple schema formats.

● Built machine learning models using Scikit-Learn and Spark MLlib for recommendation systems and ad targeting algorithms.

● Applied Natural Language Processing techniques to extract product attributes and keywords from unstructured product descriptions.

● Deployed data processing pipelines and ML services using cloud platforms including AWS, Azure, and GCP.

● Containerized backend services using Docker and deployed scalable microservices using Kubernetes-based infrastructure.

● Collaborated with cross-functional teams to design scalable backend architecture supporting high-volume advertising workloads.

Skills

Programming Languages

Python, SQL, Java, JavaScript, Bash

AI / ML Frameworks

PyTorch, TensorFlow, Keras, Scikit-Learn, XGBoost, LightGBM, Spark MLlib

Backend Development

Python Software Development, RESTful API

Development, Microservices Architecture

Cloud Platforms

AWS, Microsoft Azure, Google Cloud Platform (GCP)

Containerization & Orchestration

Docker, Kubernetes

Development Tools

Git, GitHub, GitLab, Jenkins, Jupyter Notebook,

VS Code

Machine Learning & Data Science

Machine Learning, Deep Learning, Supervised

& Unsupervised Learning, Reinforcement

Learning, Feature Engineering, Model

Evaluation

Natural Language Processing & LLM

LLM, NLP, Transformer Models,

Conversational AI, Chatbot Development,

Semantic Search, Hugging Face

Transformers, spaCy, NLTK, Gensim,

OpenNLP

MLOps & ML Infrastructure

MLflow, Kubeflow, Apache Airflow, Model

Deployment, Experiment Tracking

Big Data Technologies

Apache Spark, Hadoop, Apache Kafka, ETL

Pipelines

Education

Bachelor of Engineering - BE, Computer Science 2011 – 2015 University of Michigan

Contact this candidate