Firas Obeid Email & Phone Number

Ann Arbor, Mi, Us

Chicago, Il, Us

Chicago, Il, Us

Data Science & Modelling Credit Risk Team- Developing and researching new mythologies for model explainabilty frameworks- Developing tools for our data science and strategy team for robust model development, data granular profiling and model monitoring:Open sourced the following data science tools: https://huggingface.co/spaces/firobeid/Feature_Profilling_BiVarChartshttps://huggingface.co/spaces/firobeid/ML_Model_Metrics(Limited commercial license)- Developed modelling teams' toolbox scripts/dashboards: feature and model monitors, drift testing using adversarial validation, feature profiling our target variable in equal populations and equal intervals.- Developed our feature selection algorithm ( an iterative process with two pieces of logic pertaining the ML platforms feature impact calculation)- Model full life cycle: development, testing and monitoring

Tempe, Az, Us

Lead the first ever Artificial Intelligence bootcamp at ASU tailored towards equipping working professionals with Python, Unsupervised and Supervised Machine Learning, Transformers, Natural Language Processing, AI Applications, Time Series Forecasting, and AI & Data Ethics.Lead students to develop AI applications such as EDA agents, mps/video meeting summarizer, medical detection applications:- Meeting Summarizer: https://ai-project-3-meeting-summarizer-rttvfhl39dwghbbhc4ywan.streamlit.app/- EDA Agent: https://cgdproject3.streamlit.app/- Multi-Lingual Skin Cancer Detection: https://huggingface.co/spaces/kkruel/skin_cancer_detection_ai

Arlington, Virginia, Us

800+ 1-on-1 sessions with industry professionals all over US and Australia to deliver data science, visualization and financial technology tools for industry application and career development. Led students to deliver MVP and prototypes projects deployed on cloud based services, for example:- https://house-price-wizard.herokuapp.com/- https://nedalprofile.herokuapp.com/- https://share.streamlit.io/sjufan84/esg_clustering/main/streamlit_main.py- https://covidappproject.herokuapp.com/- https://www.youtube.com/watch?v=Krq2-CeDmzA

Arlington, Virginia, Us

- Financial Fundamentals in Time-Series Analysis, Financial Ratios, and Financial Analysis- Python 3, Pandas, MAtplotlib, API Interactions, and JupyterLab- Financial libraries and Tools like NumPy, SciPy, and Quantopian- Execution Algorithms, Monte Carlo Simulations, Risk-Data Aggregation, Forecasting, Financial Modeling, Modern Portfolio Theory, and Logistic Regression- Machine Learning using Algorithmic Trading, Random Forests, k-Nearest Neighbor (kNN), Support Z Vector Machines (SVM), Linear Regression, and Scikit-learn- Blockchain and Cryptocurrency experience in Solidity, Ethereum, Consensus Algorithms, Transactions, Validation, Distributed Ledger, Bitcoin, Mining and dApps.

Arlington, Virginia, Us

- Data science, latest technologies and visualization bootcamp - Advanced Statistics with Data Analytics with Python, Numpy, Pandas, and Matplotlib- SQL/NoSQL and Database management (ETL)- Machine Learning- Big Data- R- Web technologies and Scraping (JavaScript-html-flask frameworks)

Plano, Texas, Us

My previous experience in 2020 in developing my first language model (10 million parameters) from pretraining to fine tuning for an event driven trading project (found in project section of my profile), has led me to take on the development of an LLM in the educational and training spectrum at HigherU.At HigherU, we are developing a pretrained foundational model (a learning as a service as per say) that is adapted from another open source existing LLM (for a warm start point) but we are adapting our LLM to a domain specific universe, which is education, learning and training. Our LLM objective, is to learn to reason in the educational/learning/training universe of tokens and be able to reason in the linguistics of an educator/trainer. This facilitates fine tuning on downstream tasks (i.e Q/A, content generation) and even finetuning in specialized subjects (i.e Data Science, onboarding an employee, physics class, corporate training...)

Berkeley, Ca, Us

Lead 3 x FinTech Boot Camps for working professionals, simultaneously, at UC Berkeley that take a multidisciplinary approach to finance, fundamental programming, data analysis, financial machine learning and modern tools in cryptocurrency and blockchain.Curriculum: https://bootcamp.berkeley.edu/fintech/curriculum/Personal Web App/Lectures: https://firobeid.github.io/compose-plots/script.htmlTeaching Innovations: https://github.com/firobeid/Educator_Conference_2023/tree/main-The University of California, Berkeley, is accredited by the Western Association of Schools and Colleges (WASC). UC Berkeley Extension—like all other UC Berkeley schools, colleges and departments—is accredited by WASC through the University.

Gb

Experienced Associate in the US Risk Analytics and Modeling team, applying quantitative, modeling and analytical skills in a real world environment to develop or validate statistical and AI/machine learning models in the areas of credit risk, market risk, assets and liabilities management, fraud detection, anti money laundering and other functional modeling and analytics area.● Deployed on financial derivatives valuations, model review and python parsing script development throughout my time with the analytics team.

Dallas, Texas, Us

● Researched Deep learning applications for NLP to develop language model that regenerates financial news headlines for transfer learning. ● Built/developed byte level language model (10 mullion parameters) using semi supervised learning of self trained embeddings for multi-time horizon news classification on EC2 using Keras APIResults: https://github.com/firobeid/DeepLearningProjects/tree/main/LanguageModels_NewsHeadlines(USequities)● Developed and maintained python automation scripts to update PostgreSQL on aws cloud for the start up's asset prices update from various APIs (i.e Refinitiv Eikon)Technologies and Tools that where acquired:Tensorflow - Language modelling - AWS(S3-EC2) - Hierarchical Data Format version 5 (HDF5) - PostgresSQL - Eikons' Python API - LSTM - Distributed learning

Beirut, Lb

Registered treasury dealer (Market Maker) in the financial markets association (ACI-Lebanon). - Managed Banks’s Dealing Room (Trading Desk) activities that served 20 branches , seasonally, when chief dealer was out of the office.Results: Maintained on average a daily profit of $10,000 from FX and FX derivatives deals. - Placed and executed FX & FX derivatives (mainly currency swaps and forwards), futures(Euro, Spot & Mini gold, silver contracts), money/capital markets trades, overnight loans at prime rates- Authorized and managed paper cash in vaults transactions with on avg $1mio daily allocations.- Maintained banks reserve requirements with the Central Bank

Beirut, Lb

Treasury Department / Dealing Room / Cash CenterPlaced and executed FX & FX derrivatives, futures, money and capital markets deals.Authorized and managed cash in vaults transactions.

Beirut, Lb

Back Office Treasury Department-Checked, data entry and followed up on daily transactions (FX, Capital & Money Market).-Reconciled custodian and customer portfolios.-Booked and recorded daily transactions executed by the dealing room.-Sent swift messaging to counterparts to confirm and settle deals.

Beirut, Lb

-Competed and represented Lebanon and the Lebanese American University at Las Vegas in the WBFF 2014 fitness world championships -Proposed, Initiated and Organized two Annual Strongest Athlete events on campus for students and alumnus (i.e attached images)-Organized a Seminar on the impact of exercising on academics performance keynoted by a bodybuilding world champion-Supervised , part time, the Night Shift at LAU Fitness Center -Ensured safety of Students/ Staff /Faculty

Master Of Science - Ms, Financial Engineering & Risk Management

Bachelor'S Degree, Banking, Corporate, Finance, And Securities Law

Professional Diploma, Health And Physical Education/Fitness