Analytics / ML Engineer

Technical Skills

Programming Languages: Python, Pandas, Numpy, SQL

Databases: MySQL, ClickhouseDb, PostgreSql, DuckDb, Neo4j, Redis, Redshift

Frameworks: Jupyerhub, Kafka, Airflow, Spark, Dbt, Docker, AWS, Grafana

Visualization Tools: Metabase, Tableau, Qlik

Work Experience

Analytics Engineer @ 360 Digital Starter, Berlin (_April 2024 - _)s

• Developing reports and alerts for general Marketing metrics (CTR, ROI, ROAS, CPA). Setting up A/B tests for various campaigns and across multiple products.
• Setting up real-time ETL pipelines from scratch using Kafka, ClickhouseDb, Metabase and monitoring of real-time services using Docker, Grafana.
• Built and implemented a multi-touch attribution model to accurately attribute conversion across channels and optimize marketing performance insights.
• Alerts to monitor issues in deployments, downtimes, SLAs.

Data Science Werkstudent @ Sportec Solutions AG, Munich (October 2022 - November 2023)

• Synchronization Algorithm - Improve shot identification by ~5% overall accuracy over the baseline.
• Masters Thesis - Completed Research and Development for the topic “Estimating effectiveness of defensive KPIs in football.”

Machine Learning Engineer @ Simpl, Bangalore (January 2019 - October 2022)

• Developed the infra and built multiple anti-fraud predictive Machine Learning models. Overall, increased the F-1 score of the entire Anti Fraud Risk engine by ~ 41%
• Reduce p99 metrics response time for all services in the anti fraud by an average 22%.
• Developed an end-to-end automated blocking and flagging system of suspicious users.

Education

• M.Sc., Web & Data Science - University of Koblenz (November 2023)
• B.Sc., Computer Science - NIIT University (August 2019)

Projects

Data-Driven EEG Band Discovery with Decision Trees

Publication

Developed objective strategy for discovering optimal EEG bands based on signal power spectra using Python. This data-driven approach led to better characterization of the underlying power spectrum by identifying bands that outperformed the more commonly used band boundaries by a factor of two. The proposed method provides a fully automated and flexible approach to capturing key signal components and possibly discovering new indices of brain activity.

Decoding Physical and Cognitive Impacts of Particulate Matter Concentrations at Ultra-Fine Scales

Publication

Used Matlab to train over 100 machine learning models which estimated particulate matter concentrations based on a suite of over 300 biometric variables. We found biometric variables can be used to accurately estimate particulate matter concentrations at ultra-fine spatial scales with high fidelity (r2 = 0.91) and that smaller particles are better estimated than larger ones. Inferring environmental conditions solely from biometric measurements allows us to disentangle key interactions between the environment and the body.

Talks & Lectures

• Causality: The new science of an old question - GSP Seminar, Fall 2021
• Guest Lecture: Dimensionality Reduction - Big Data and Machine Learning for Scientific Discovery (PHYS 5336), Spring 2021
• Guest Lecture: Fourier and Wavelet Transforms - Scientific Computing (PHYS 5315), Fall 2020
• A Brief Introduction to Optimization - GSP Seminar, Fall 2019
• Weeks of Welcome Poster Competition - UTD, Fall 2019

• A Brief Introduction to Networks - GSP Seminar, Spring 2019

• Data Science YouTube

Publications

1. Talebi S., Lary D.J., Wijeratne L. OH., and Lary, T. Modeling Autonomic Pupillary Responses from External Stimuli Using Machine Learning (2019). DOI: 10.26717/BJSTR.2019.20.003446
2. Wijeratne, L.O.; Kiv, D.R.; Aker, A.R.; Talebi, S.; Lary, D.J. Using Machine Learning for the Calibration of Airborne Particulate Sensors. Sensors 2020, 20, 99.
3. Lary, D.J.; Schaefer, D.; Waczak, J.; Aker, A.; Barbosa, A.; Wijeratne, L.O.H.; Talebi, S.; Fernando, B.; Sadler, J.; Lary, T.; Lary, M.D. Autonomous Learning of New Environments with a Robotic Team Employing Hyper-Spectral Remote Sensing, Comprehensive In-Situ Sensing and Machine Learning. Sensors 2021, 21, 2240. https://doi.org/10.3390/s21062240
4. Zhang, Y.; Wijeratne, L.O.H.; Talebi, S.; Lary, D.J. Machine Learning for Light Sensor Calibration. Sensors 2021, 21, 6259. https://doi.org/10.3390/s21186259
5. Talebi, S.; Waczak, J.; Fernando, B.; Sridhar, A.; Lary, D.J. Data-Driven EEG Band Discovery with Decision Trees. Preprints 2022, 2022030145 (doi: 10.20944/preprints202203.0145.v1).
6. Fernando, B.A.; Sridhar, A.; Talebi, S.; Waczak, J.; Lary, D.J. Unsupervised Blink Detection Using Eye Aspect Ratio Values. Preprints 2022, 2022030200 (doi: 10.20944/preprints202203.0200.v1).
7. Talebi, S. et al. Decoding Physical and Cognitive Impacts of PM Concentrations at Ultra-fine Scales, 29 March 2022, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-1499191/v1]
8. Lary, D.J. et al. (2022). Machine Learning, Big Data, and Spatial Tools: A Combination to Reveal Complex Facts That Impact Environmental Health. In: Faruque, F.S. (eds) Geospatial Technology for Human Well-Being and Health. Springer, Cham. https://doi.org/10.1007/978-3-030-71377-5_12
9. Wijerante, L.O.H. et al. (2022). Advancement in Airborne Particulate Estimation Using Machine Learning. In: Faruque, F.S. (eds) Geospatial Technology for Human Well-Being and Health. Springer, Cham. https://doi.org/10.1007/978-3-030-71377-5_13

• Data Science Blog