Git Product home page Git Product logo

👋 Hello, I'm Ephrem Tadesse Degu!

Objective

Passionate data scientist with a robust background in credit scoring, machine learning, and natural language processing (NLP), aiming to leverage advanced analytics to drive business insights, optimize decision-making processes, and foster innovation in dynamic environments.

Technical Skills

  • Programming Languages: Python, Java, Shell Scripting
  • Machine Learning/MLOps: Scikit-learn, TensorFlow, MLflow, Kubeflow
  • Data Engineering: Kedro, Redash, Pandas Profiling, Seaborn, NumPy
  • Web Frameworks: FastAPI, Flask
  • Cloud Services: AWS (S3, Lambda, EC2, CloudFormation, CloudWatch, RDS, AWS-CLI, EKS, EBS, EFS, ECR, IAM roles and management)
  • Version Control: GitHub, GitLab
  • Containerization & Orchestration: Docker, Kubernetes/EKS
  • CI/CD & Deployment: Terraform, ArgoCD, GitOps
  • Project Management: Jira

Projects

Credit Scoring Engine for a Digital Lending Platform

  • Tools/Frameworks: AWS Services, FastAPI, Docker, MLflow, Kubeflow, feast
  • What it does: Developed an alternative data-driven credit scoring and configurable decisioning platform for short-tenure loan products. Implemented and deployed multiple machine learning models, establishing fully automated MLOps pipelines for seamless integration and operational efficiency.
  • Evaluation of stability: Monitored accuracy, performance metrics, and robustness to data shifts and noise. Successfully processed over 1,000,000 requests with response times consistently under 1 second, serving a user base of 60,000 unique customers over a one-year period.

Implementation of Datalake and Analytics Tool for Financial Report Data

  • Tools/Frameworks: Redash, AWS Services (S3, EC2, RDS, Lambda), Kedro
  • What it does: Developed an ETL pipeline for an edutech company to extract over six years of financial report data and integrate it with their ledger software. Built analytics dashboards to provide actionable insights into financial activities.
  • Usability: Enabled a 30% increase in user adoption of financial analytics dashboards and reports, leading to a 25% reduction in time spent on data retrieval and analysis.

End-to-End Monitoring and Evaluation (M & E) KPI Metadata Mapping Data Pipeline

  • Tools/Frameworks: AWS Services (S3, EC2, RDS, Lambda), Kedro
  • What it does: Developed an ETL pipeline for an edutech company to extract their 3-year OKR planning data and integrate it with available resources. Implemented semantic keyphrase embedding and graph techniques to resolve hierarchical-based conceptual data mapping and grouping challenges. Built analytics dashboards for actionable insights into KPI and OKR mapping.
  • Impact: Enabled a 15% increase in user adoption of KPI and OKR mapping analytics dashboards and reports, with a 10% reduction in time spent on monitoring progress against objectives.

Personalized Document Support Conversational Bot

  • Tools/Frameworks: Langchain, FastAPI, JavaScript
  • What it does: Integrated OpenAI GPT-3.5 API for document-based question answering, improving accessibility and usability of financial reports and analytics insights. Utilized Langchain with FastAPI and JavaScript to enable natural language interaction, simplifying complex financial data for actionable insights.

Event and Temporal Information Extraction from Amharic Text

  • Tools/Frameworks: Python, LSTM, TensorFlow
  • What it does: Developed methods to extract event and temporal information from Amharic text. Constructed a dataset of unstructured temporal-triggered news headlines in Amharic. Addressed challenges in disambiguating deverbal nominal entities through rule-based approaches and classical ML models.

Experience

  • Senior Data Scientist at Kifiya Financial Technologies
    (Led credit scoring initiatives and implemented MLOps practices for digital lending platforms)

  • Data Scientist at Tenacious Intelligence Corporation
    (Led development of scoring engines and created business analytics dashboards)

  • Text Data Analytics Team Leader at Ethiopian Artificial Intelligence Institute, Ethiopia
    (Led projects in NLP and implemented data pipelines for machine translation)

  • Lecturer at Jimma University
    (Developed data science curriculum and conducted research in Natural Language Processing)

Contact

Ephrem Tadesse's Projects

amhariccorpus icon amhariccorpus

The set of files used for the development of the Amharic Corpus.

data icon data

Lexical Data of Ge'ez Languages

data-science-ipython-notebooks icon data-science-ipython-notebooks

Data science Python notebooks: Deep learning (TensorFlow, Theano, Caffe, Keras), scikit-learn, Kaggle, big data (Spark, Hadoop MapReduce, HDFS), matplotlib, pandas, NumPy, SciPy, Python essentials, AWS, and various command lines.

datasets icon datasets

🤗 The largest hub of ready-to-use datasets for ML models with fast, easy-to-use and efficient data manipulation tools

ethiotelecomcdranalysis icon ethiotelecomcdranalysis

Ethiotelecom is one of the giant network provider company located in Ethiopia. Due to increasing demands and infrastructure limitation the government has decided to outsource Ethiotelecom for additional network providers. Following this expansion, the company needs an intensive research on mobile pattern traffic analysis, spatiotemporal analysis of CDR (Call Detail Record) data, temporal correlation to extract mobile traffic pattern, developing generic data-driven resource allocation approach for cellular networks based on CDR activity levels etc. Motivated by this, we perform Exploratory analysis and prediction selected features of CDR data gained from Ethiotelecom. Thus, on the basis of temporal insights of total call duration, call fee and network download traffic, a framework has been proposed for mobile traffic pattern clustering. Moreover, timeseries analysis and forecasting of CDR features will be conducted soon.

personalizedgptpoweredqa icon personalizedgptpoweredqa

This repo enables you to simplify your tasks to get relevant and precise ifnormation from your documents. You can upload your documents or provide the directories where your list of documents found.

simpleamharicocr icon simpleamharicocr

This repository is simple Amharic printed document character recognition using open source tools and streamlit. Further, this application is deployed on heroku development serever. In the near future we will come up with handwritten and printed character recognition system for all Ethiopic scripts. Thus, if anyone wants to collaborate with our project you can ping me with email.

streamlitpredictionapp icon streamlitpredictionapp

This repository is about simple prediction model using classical supervised machine learning models. We built our model and integrate with streamlit. Finally, we deploy the simple prediction model on heroku development server.

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.