Gültekin G.
Machine Learning Management
Experience
Jan 2021 - Present
5 yearsSan Francisco, United States
Machine Learning Management
Turing.com
- Led a team of 75 Data Scientist for improving Gemini Advanced in collaboration with Google Bard’s Implicit Code Execution (ICE) team.
- Owned model evaluation and error discovery processes. Generated metrics on model correctness and compared the various model versions to achieve a steady improvement. For statistically significant error patterns, reasons were investigated and fix proposals were discussed with the research teams.
- Recommended and implemented new policies to the model in terms of generating training data to be consumed by SFT processes.
- Led both high quality throughput RLHF training data generation to further improve the model.
- Worked on addition of RLMF processes to refine the model using synthetically generated data.
- Contributed to the establishment of the Machine Learning (ML) Science and Data Analysis foundations + team. Owned various model training -classifiers and recommendation systems- and deployment, the creation of robust validation frameworks to ensure our ML models’ performance is future-proof, and continuous feature engineering to enhance our existing recommendation models, directly improving business metrics.
- integrated new signals into our models, such as Latent Dirichlet Allocation (LDA) based feature extraction from unstructured text data.
- For ML Engineering utilized Docker, ElasticSearch, Google Cloud Platform (VertexAI, BigQuery, GCS), and Uvicorn. While also used ML Frameworks such as Tensorflow, Pytorch, LightGBM, XGBoost, Huggingface, Scikit-Learn, Imblearn, LightFM, and OpenAI API.
- In my leadership role, I have effectively managed a team of 7 Data Scientists and fostered active collaboration with cross-functional teams across the organization.
Aug 2019 - Feb 2021
1 year 7 monthsSenior ML Scientist
Omnisight
- Developed deep neural networks for outdoor advertisement insights, including human detection, face recognition, and pose estimation.
- Enhanced model performance on NVIDIA Jetson with TensorRT optimization, increasing throughput and efficiency.
- Deployed a multi-object tracking system to deliver real-time analytics, elevating customer experience.
Sep 2018 - Mar 2019
7 monthsML Scientist
Science Wave Capital
- Created predictive models for forecasting stock returns and 5-day alpha values for a market-neutral portfolio using LightGBM, XGBoost, and neural networks on 20 years of European stock data.
- Conducted feature engineering to develop new signals for model adaptation to volatile market conditions.
- Facilitated daily reallocation of a $150M leveraged portfolio, integrating model predictions with Markowitz optimizers for optimal results.
Nov 2016 - Sep 2018
1 year 11 monthsAI Engineer
Infotech
- Led AI efforts in the TEMSA MD9 electriCITY autonomous bus project, focusing on sensor data processing and fusion from LIDAR/Stereo Cameras for computer vision tasks such as object detection and tracking.
- Implemented machine learning models for government RD, including indoor localization using smartphones and driver risk classification using telematics.
Languages
Turkish
NativeEnglish
IntermediateEducation
Oct 2015 - Jun 2017
Galatasaray University
Master of Science, Computer Science & Engineering · Computer Science & Engineering · İstanbul, Turkey
Oct 2009 - Jun 2014
Sabanci University
Bachelor of Science, Computer Science & Engineering · Computer Science & Engineering · İstanbul, Turkey
Oct 2005 - Jun 2009
Izmir American College
High School Diploma · İzmir, Turkey
Need a freelancer? Find your match in seconds.
Try FRATCH GPT More actions
Similar Freelancers
Discover other experts with similar qualifications and experience
- English
- Deutsch
2025 © FRATCH.IO GmbH. All rights reserved.