Mustafa Çağlar, Developer in London, United Kingdom
Mustafa is available for hire
Hire Mustafa

Mustafa Çağlar

Verified Expert  in Engineering

Data Scientist and Developer

Location
London, United Kingdom
Toptal Member Since
January 24, 2022

With a PhD in physics from Cambridge University, Mustafa has over five years of data science experience, including three years in fast-paced startups. His published work showcases exceptional visualization skills. An expert in applying AI to business challenges, he has collaborated with military organizations, NGOs, and fintech and civil service clients. Specializing in deep learning and computer vision, Mustafa adopts an iterative Agile workflow for rapid development.

Portfolio

Faculty
Computer Vision, Natural Language Processing (NLP), AWS Deployment, Python...
Faethm
Amazon Web Services (AWS), Time Series Analysis, Forecasting, Python...
Tymit
Python, Amazon Web Services (AWS), Amazon SageMaker, Regression, Classification...

Experience

Availability

Part-time

Preferred Environment

PyCharm, Visual Studio Code (VS Code), Python, Amazon Web Services (AWS), MacOS, Jupyter Notebook

The most amazing...

...thing I've achieved is revolutionizing a document digitization pipeline for the entire UK, leveraging AI to enable a remarkable 50% manual effort reduction.

Work Experience

Senior Data Science Consultant

2022 - PRESENT
Faculty
  • Developed an AI-enabled digital twin of a classified organization. This was joined with a deployment in AWS and a design and deployment of UI features.
  • Created and deployed zero-shot computer vision techniques for document question answering on scanned documents for a large governmental organization.
  • Conducted real-time data ingestion from social networks with media capture and real-time OCR. Used a large language model (LLM) to classify, summarize, and topic analyze.
Technologies: Computer Vision, Natural Language Processing (NLP), AWS Deployment, Python, Machine Learning, Pandas, Deep Learning, Artificial Intelligence (AI), Architecture

Senior Data Scientist

2022 - 2022
Faethm
  • Mined data across a variety of sources using NLP and warehousing.
  • Forecasted medium-long-term trends in workforce behaviors.
  • Communicated with stakeholders and clients via a proprietary SaaS platform.
Technologies: Amazon Web Services (AWS), Time Series Analysis, Forecasting, Python, Data Science, Agile, Looker, SQL, Data Scraping, Natural Language Processing (NLP), Machine Learning, Deep Learning

Full-stack Data Scientist

2019 - 2022
Tymit
  • Created the entire analytics function moving the company from SharePoint and Excel to AuroraDB and BI tools, Looker and Metabase.
  • Developed and deployed a credit risk model to reduce financial loss and increase credit approvals.
  • Implemented a pipeline to process user survey responses.
  • Optimized conversion rate via social network advertisements.
  • Implemented and improved facial recognition services atop AWS Rekognition.
Technologies: Python, Amazon Web Services (AWS), Amazon SageMaker, Regression, Classification, Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), GPT, Predictive Modeling, Predictive Analytics, Data Warehousing, Data Science, SQL, Data Analytics, Machine Learning, Pandas, Deep Learning

Data Scientist | Visiting Fellow

2020 - 2020
The Francis Crick Institute
  • Developed a completely novel approach allowing live-cell classification.
  • Directed impact to shaping how future experiments will be designed for Alzheimer's cell classification.
  • Generated real insight into Parkinson’s drug efficacy by way of model interpretability.
Technologies: Python, TensorFlow, SQL, Machine Learning

Researcher | Postgraduate

2015 - 2020
University of Cambridge
  • Created fully automated data pipelines to perform live analysis and machine learning on data streams, warehoused in Azure.
  • Analyzed large datasets with complex regression functions in MATLAB and R in order to demonstrate novel physical concepts.
  • Contributed and authored three high-impact journal papers showcasing visualizations created in Python and insights garnered using machine learning.
  • Taught laboratory practicals in C++ and Python to undergraduates at the University of Cambridge; overall, four cohorts of 20 students.
  • Supervised and mentored four master's students through year-long projects with topics ranging from computer vision to complex classification.
Technologies: Python, Machine Learning, Azure, Regression, Classification, ImageJ, Machine Vision, Data Science, SQL

Consultant Engineer

2015 - 2015
Sagentia
  • Developed an algorithm in MATLAB for measuring curliness in hair from a photograph and translated it to OpenCV implementation on iOS and ARM devices.
  • Created a native iOS app, published to the AppStore, to communicate and monitor implanted pacemakers.
  • Worked on internal data collection and analytics using Python to streamline project budgets by optimizing project assignments.
Technologies: C++, Python, Analysis, iOS, MATLAB, OpenCV

Scorecard for Credit Worthiness

https://tymit.com
A high dimensional model developed in Python to consume data from credit bureaus, such as Experian, combined with mined behavioral features to provide a risk of default score to each applicant. The model is tuned according to the company's KPIs, deployed, and currently in use in production.

Atomic Image Reconstruction Using Neural Networks

A Python module created to construct high-resolution atomic imagining from electron wave functions. A trained neural network is used to fit as closely as possible to wave functions. The subsequent functions are using alongside PyQSTEM to reconstruct very high-resolution images.

It was published in ACS Nano.

Detecting DNA Translocation Signal

https://iopscience.iop.org/article/10.1088/1361-6463/abe07b
DNA strands moving through a hole and pore can create electrical signals that can be used to sequence the base pairs present. However, these signals are weak and significantly weaker than the overpowering random noise. Signal analysis techniques and clustering algorithms are used to unpick signals from noise and identify different lengths and composition of DNA present in samples.

It was published in IOP.

Languages

Python, SQL, C++, C, Verilog, R

Libraries/APIs

Pandas, TensorFlow, Keras, OpenCV

Tools

PyCharm, MATLAB, Amazon SageMaker, Looker, ImageJ, AWS Deployment

Paradigms

Data Science, Agile

Platforms

MacOS, Visual Studio Code (VS Code), Amazon Web Services (AWS), Jupyter Notebook, Azure, iOS

Other

Research, Technical Writing, Regression, Classification, Predictive Modeling, Predictive Analytics, Machine Learning, Artificial Intelligence (AI), Publishing, Natural Language Processing (NLP), Time Series Analysis, Forecasting, Data Warehousing, Data Scraping, GPT, Generative Pre-trained Transformers (GPT), Data Analytics, Deep Learning, Architecture, Signal Processing, Digital Signal Processing, Computer Vision, Clustering, Machine Vision, Analysis

2016 - 2019

Doctorate Degree in Physics

University of Cambridge - Cambridge, UK

2015 - 2016

Master's Degree in Electrical Engineering

University of Cambridge - Cambridge, UK

2011 - 2015

Undergraduate Master's Degree in Electrical Engineering

University of Southampton - Southampton, UK

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.

1

Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring