Ahmet Hamza Emra
Verified Expert in Engineering
Machine Learning Engineer and Software Developer
Vancouver, BC, Canada
Toptal member since July 10, 2020
Ahmet is a skilled machine learning engineer with expertise in deep learning and reinforcement learning. He has worked on projects in academia and industry in a variety of team and client settings. One of his most exciting projects was in collaboration with NASA. Ahmet is currently working on research in advanced reinforcement learning.
Portfolio
Experience
- Machine Learning - 7 years
- Python - 6 years
- Artificial Intelligence (AI) - 6 years
- SQL - 5 years
- Data Science - 4 years
- Deep Learning - 4 years
- TensorFlow - 4 years
- Reinforcement Learning - 3 years
Availability
Preferred Environment
Reinforcement Learning, Machine Learning, Artificial Intelligence (AI)
The most amazing...
...product I've developed was an analytical tool to help astronauts maintain a healthy environment in the International Space Station.
Work Experience
Lead AI Engineer
The Vocational Rehabilitation Technical Assistance Center for Quality Management
- Led the design and implementation of an advanced retrieval-augmented generation (RAG) chatbot for a vocational rehabilitation platform, integrating NLP models with vector similarity searches to deliver personalized, context-aware assistance to users.
- Developed a two-step query pipeline with dynamic search term generation and relevance-based context filtering, enhancing accuracy by 20% while efficiently managing large-scale documents and real-time data retrieval, increasing token usage by 12%.
- Implemented advanced conversation management using Redis for scalable session handling and designed custom prompts with domain-specific guidelines.
Senior Data Scientist
Renegade Publishing Inc.
- Built an ETL pipeline to detect the continuity of royalty payments.
- Integrated Apache Airflow into the current system and hosted it on AWS.
- Worked to predict royalty payments and leakage to improve reporting and analysis pages.
LangChain Developer
US Dream Academy
- Developed a chatbot with an RAG system to ensure answer quality and confidence.
- Implemented a document ranking model to help document retrieval.
- Developed and deployed the chatbot UI to Azure web services.
Lead ML Developer
DigitalVibesCA
- Developed a custom chatbot that utilizes a custom GPT and is supported with a new Flask API to retrieve related data.
- Enhanced the RAG system by leveraging open-source LLM to transform tabular data into a more suitable format for ingestion by GPT-4.
- Headed a team of developers and mentored them on system architecture.
Lead AI Developer (via Toptal)
Proxa
- Developed and implemented AI models to enhance data management and reporting functionalities, contributing to the advancement of an AI-enhanced platform.
- Utilized extensive experience with Microsoft Azure and demonstrated expertise in leveraging and creating generative AI applications to support clients' needs using LLMs and LangChain technologies.
- Architected the data hub and formulated contextualized models based on data, optimizing the system's capabilities and performance.
Machine Learning Expert
The Listening App Inc.
- Implemented a TensorRT version of Bark, an open source text-to-audio model, enhancing the model's speed by 35%.
- Introduced Proximal Policy Optimization (PPO) to fine-tune the Bark model using RLHF.
- Came up with a reward signal for reinforcement learning training.
Data Scientist and Consultant
MovieUnion
- Trained a recommendation system algorithm from scratch using TensorFlow.
- Designed databases, stored procedures, reports, and data input interfaces using SQL.
- Deployed the machine learning model to SageMaker.
Data Scientist
Pfizer
- Spearheaded the development and training of predictive models to optimize drug production and inspection timelines, reducing forecasting errors from eight days to an average of two days and significantly enhancing operational efficiency.
- Architected and implemented scalable ETL pipelines using Python, PostgreSQL, and Apache Airflow, streamlining data processing workflows and improving data integration across the organization.
- Led the modernization of a legacy data storage system, designing a new version to enhance data retrieval, updating capabilities, and scalability, driving improved system performance and user experience.
Data Scientist
InkSpace Analytics - Freelance
- Designed process-oriented workflow and ELT processes for loading data into the data warehouse.
- Trained a predictive model on loan approval and amount using TensorFlow.
- Worked closely with project managers, business analysts, source owners, and the data quality team to ensure timely and accurate delivery of business requirements.
- Managed the performance tuning of SQL queries.
Artificial Intelligence Engineer
VUCA INC
- Developed a deep reinforcement learning agent to solve design problems, using PyTorch.
- Built the web interface of a reinforcement learning environment for training and evaluation.
- Integrated gaze point detection for model training.
Research Assistant
University of Alberta
- Researched the application of reinforcement learning algorithms in a supervised setting.
- Developed deep Q-cogent networks, a new deep reinforcement learning (DRL) algorithm that improved learning speed by 50%.
- Performed data reporting and data reconciliation processes to support the principal project investigators.
- Initiated the publication of an academic paper on DRL.
Data Analyst
Harmony Public Schools
- Developed a tool to analyze students' and teachers' interactions to improve the success of both parties.
- Provided complex research and analysis to support effective teaching and presented findings to management or project leaders.
- Created weekly and per-semester reports to provide advanced insights and suggestions to the administration.
Undergraduate Research Assistant
North American University
- Researched a data-driven approach to bacteria development with Dr. Zubeyir Altintas.
- Taught multiple classes including Python and deep learning.
- Expanded breadth and depth of knowledge and skills across disciplines by working with specific fields of computer science and industry professionals.
Experience
CO2 Analysis on the International Space Station (ISS)
https://drive.google.com/file/d/1lrRF9z63MbMlJMOyTDxOsCxewMSF4qC5/viewMLRpro
https://github.com/MichaelMMeskhi/MLRPIntelegent_Lock
https://github.com/AhmetHamzaEmra/Intelegent_LockGenerating Logos with Diffusion Models
Clone a Friend
Education
Bachelor's Degree in Computer Science
North American University - Houston, TX, USA
Certifications
Deep Learning Specialization
Coursera
Building and Deploying Deep Learning Applications with TensorFlow
Linkedin Learning
OpenCV for Python Developers
Linkedin Learning
Machine Learning Nanodegree
Udacity
Data Analysis Nanodegree
Udacity
Skills
Libraries/APIs
Pandas, NumPy, Scikit-learn, TensorFlow Deep Learning Library (TFLearn), TensorFlow, Keras, PyTorch, PySpark, D3.js, OpenCV, Spark ML, Natural Language Toolkit (NLTK), PyPDF2, Google Speech-to-Text API
Tools
Jupyter, ChatGPT, Terminal, Tableau, Slack, Spark SQL, Amazon SageMaker, Google AI Platform, Open Neural Network Exchange (ONNX), Text2Speech, Apache Airflow, AI Prompts
Languages
Python, Python 2, Python 3, SQL, HTML, CSS, R, JavaScript
Platforms
Windows, Debian, Linux, Amazon Web Services (AWS), Azure, AWS Lambda, MacOS, Docker, KNIME
Frameworks
Apache Spark, Flask, Django, Spark
Storage
Azure SQL, Amazon S3 (AWS S3), PostgreSQL, MySQL, Redshift, Data Pipelines, Azure Websites
Paradigms
ETL, Continuous Integration (CI), Continuous Development (CD), Rapid Prototyping
Other
Deep Learning, Machine Learning, Reinforcement Learning, Data Virtualization, Deep Reinforcement Learning, Neural Networks, Deep Neural Networks (DNNs), Data Wrangling, Data Cleaning, Data Science, Data Mining, Artificial Intelligence (AI), Optical Character Recognition (OCR), Statistical Analysis, Predictive Analytics, Data Engineering, Web Scraping, Natural Language Processing (NLP), Computer Vision, Supervised Learning, Generative Pre-trained Transformers (GPT), Classifier Development, Data Analysis, Image Recognition, Object Detection, Image Processing, Recommendation Systems, Text Processing, Liveness Detection, Anti-spoofing, Video Processing, Convolutional Neural Networks (CNNs), Facial Recognition, Forecasting, Amazon Machine Learning, Statistics, Statistical Learning, Models, Pharmaceuticals, Tracking, Statistical Process Control Tools, Diffusion Models, Stable Diffusion, Generative Models, Technical Consulting, Google Cloud Machine Learning, CSV, MVP Design, Text to Speech (TTS), OpenAI GPT-3 API, Generative Pre-trained Transformer 3 (GPT-3), Bark, Software, Software Architecture, OpenAI GPT-4 API, LangChain, Large Language Models (LLMs), Llama 2, Pinecone, Chatbots, Live Chat, APIs, API Integration, OpenAI, DQN, Actor-critic Methods (A2C, A3C), Games, Speech to Text, Open-source LLMs, LoRa, Hugging Face, LLama7B, GPU Computing, Multi-agent Systems, Machine Learning Operations (MLOps), Retrieval-augmented Generation (RAG), FastAPI, Prompt Engineering, ChatGPT Prompts
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring