Senior Full-Stack Data Scientist |Fintech |MLOps

I'm a senior data scientist with a strong background in credit scoring, NLP, and end-to-end machine learning systems. I currently lead the development of a no-code, AI-powered credit decisioning platform with LLM integration. I specialize in building scalable, production-grade ML and data pipelines and have successfully managed cross-functional teams to deliver high-impact, real-world solutions in fintech and beyond.

• Languages: Python, Java, Bash
• Machine Learning & MLOps: Scikit-learn, TensorFlow, MLflow, Kubeflow, Feast, Featureform,
• Data Engineering: Kedro, Redash, Pandas Profiling, NumPy, Seaborn
• Web Frameworks: FastAPI, Flask
• Cloud & DevOps: AWS (EC2, Lambda, RDS, S3, EKS, CloudWatch, IAM, ECR, EFS), Docker, Kubernetes, GitOps, Terraform, ArgoCD
• Version Control & CI/CD: GitHub, GitLab, GitHub Actions
• Project Management: Jira

Stack: FastAPI, Docker, MLflow, AWS, Kubeflow
Developed an alternative data-powered credit scoring system and configurable decision engine. Served 200K+ customers, processed 1M+ requests with sub-second latency, and supported over $200M in disbursed loans across 2 years. Monitored and optimized model performance with Grafana and MLflow.

Stack: Kedro, Redash, AWS (S3, Lambda, RDS)
Built a robust ETL pipeline to unify six years of ledger and financial data. Delivered business intelligence dashboards that improved adoption by 30% and reduced manual reporting efforts by 25%.

Stack: Kedro, AWS, NLP Graph Techniques
Created a smart semantic mapping pipeline to connect three years of strategic OKR plans to operational KPIs. Enabled automated grouping and dashboarding with a 10% reduction in evaluation time and a 15% boost in stakeholder engagement.

Stack: Langchain, FastAPI, OpenAI GPT-3.5, JavaScript
Built a natural language interface to enable users to query structured financial insights directly from uploaded documents. Simplified complex data interpretation using LLMs for decision-makers.

Stack: TensorFlow, LSTM, Rule-based NLP
Pioneered an approach for temporal information extraction from Amharic news texts. Constructed a novel dataset and applied both rule-based and ML methods to address semantic ambiguity in temporal expressions.

• Founder & Lead Data Scientist, Bahirbits Fintech Solutions
  Designing and building CredE$hi — an AI-first, no-code credit decisioning platform to empower lenders in underserved markets.

• Senior Data Scientist, Kifiya Financial Technologies
  Led credit scoring infrastructure development and deployed scalable MLOps workflows for high-volume lending platforms.

• Data Scientist, Tenacious Intelligence Corporation
  Built ML-driven scoring engines and developed business intelligence dashboards for operational insight.

• Team Lead, NLP Group, Ethiopian Artificial Intelligence Institute
  Directed NLP R&D projects, including machine translation, text classification, and corpus curation for low-resource languages.

• Lecturer, Jimma University
  Taught data science and NLP, and co-developed curriculum focused on applied machine learning and natural language technologies.

• Email: [email protected]
• (Feel free to reach out for collaboration, mentorship, or speaking opportunities.)