I'm a Generative AI Engineer, Data Scientist, and Full-Stack Consultant based in Addis Ababa, Ethiopia 🇪🇹. I design and ship production-grade AI systems — RAG pipelines, MCP servers, data platforms, and the backend architecture underneath them — for startups, scale-ups, and innovation teams across Africa and beyond.
Away from client work I operate as Mr. Fish 🦈 — a solo consulting practice built on one promise: one expert, zero hand-offs, zero excuses.
profile = {
"name": "Fisseha Estifanos",
"alias": "Mr. Fish 🦈",
"location": "Addis Ababa, Ethiopia 🇪🇹 (working globally 🌍)",
"focus": ["Generative AI & RAG", "MCP & Agentic Systems", "Data Engineering", "Full-Stack"],
"education": ["BSc Software Engineering — AAiT", "10 Academy Data Engineering Fellowship"],
"currently": "Sr. Data Science & Analytics Consultant @ Kleene.ai",
"open_to": ["Consulting engagements", "Open-source collaboration", "Advisory roles"],
"fun_fact": "Writes about Stoicism between deployments 📚",
}- 📊 Senior Data Science & Analytics Consultant @ Kleene.ai — turning raw business data into decision-grade analytics on the kleene platform
- 🦈 Mr. Fish — my independent practice: RAG architectures, technical audits, rapid MVPs, and fractional-CTO work
- 🔬 Exploring — Model Context Protocol (mcp_explore), agentic workflows, and evaluation-driven LLM development
Generative AI & LLMs
Data Engineering & Machine Learning
Backend & Full-Stack
Cloud, DevOps & Web3
A curated slice of public work — most client engagements live in private repositories.
| Project | What it does | Stack |
|---|---|---|
| LLM-API ⭐ | Large language models in practice — in-context learning and prompt engineering with open-source LLMs (Bloom, Cohere) | Python · LLMs · Prompt Engineering |
| mcp_explore | Hands-on exploration of the Model Context Protocol for tool-using AI agents | Python · MCP |
| ELT | End-to-end ELT pipeline: Airflow orchestration, dbt transformations, data-quality gates, Redash dashboards | Airflow · dbt · PostgreSQL · Great Expectations |
| Rossmann-pharmaceuticals | Sales forecasting six weeks ahead across hundreds of stores with ML and LSTM deep learning | Scikit-Learn · TensorFlow · DVC · MLflow |
| delivery_drivers | Causal inference + logistic optimisation to position Gokada delivery drivers in Lagos | Python · Causal Inference · Geospatial |
| telecom / visualizer | Telecom user analytics from raw ingestion to a self-serve Streamlit dashboard | Pandas · Streamlit · PostgreSQL |
| twitter-visualizer | Sentiment analysis and topic modelling of ~22k tweets, served as an interactive dashboard | NLP · LDA · Streamlit |
| algorand_dApp | Web3 dApp minting and distributing NFT certificates on the Algorand blockchain | Algorand · Smart Contracts · Flask |
| defund_dApp | Location-aware smart-contract escrow — funds release only when GPS conditions are met | Solidity · Web3 |
Open-source & team contributions: Arclio/arclio-mcp-tooling (MCP development tooling) · TenAcademy/backtesting (crypto & stock backtesting infrastructure) · TenAcademy/text-to-speech data collection (Kafka/Spark/Airflow pipeline for Amharic & Swahili speech data)
I write on Medium about data science, Web3, and the occasional dose of philosophy:
- 📈 Twitter Data Analysis — sentiment analysis & topic modelling · Aug 2022 — an end-to-end NLP pipeline over ~22,000 tweets
- ⛓️ Web3 and the Algorand Blockchain Network · Oct 2022 — building an NFT-certificate dApp, from Web1 history to proof-of-stake
- 🏛️ Stillness Is the Key · Aug 2023 — Seneca, focus, and staying calm amid chaos
| Period | Organisation | Role |
|---|---|---|
| 2026 – Present | Kleene.ai | Senior Data Science & Analytics Consultant |
| 2025 – 2026 | Arclio | Generative AI Engineer — MCP tooling & AI agents |
| 2022 – Present | Mr. Fish Consulting | Founder & Principal Consultant — GenAI · RAG · Data · Full-Stack |
| 2022 | 10 Academy — Batch VI | Data Engineering Fellow — five production-grade projects |
| 2019 – 2021 | iCare Ethio Medical | Backend Developer & Database Engineer |
| 2019 | iCare Ethio Medical | Software Engineering Intern |
Highlights — the numbers behind the roles
- iCare Ethio Medical — built the C# / ASP.NET API and MSSQL data layer for a hospital management platform; cut patient waiting time from 30+ minutes to under 2 in the lobby, laboratory, and pharmacy wings, and from 20+ minutes to under 5 in doctors' rooms.
- 10 Academy — completed an intensive data-engineering fellowship: five production-grade projects in 16 weeks spanning ML forecasting, causal inference, ELT warehousing, Kafka streaming, and Web3.
- Mr. Fish — 10+ projects shipped for clients across 5+ industries in North America, Europe, Asia-Pacific, and the Middle East.
| Credential | Institution | Year |
|---|---|---|
| BSc, Software Engineering | Addis Ababa Institute of Technology (AAiT) | 2018 – 2022 |
| Data Engineering Fellowship | 10 Academy | 2022 |
| Software Engineering Programme | ALX | 2022 |
| Deep Learning Course | Global AI Hub | 2021 |
I take on a small number of consulting engagements at a time so every project gets full senior attention. If you're building something that needs serious AI integration, data engineering, or backend architecture — let's talk.
"Not theory. Not slide decks. Track record."