Based in Nairobi, Kenya · Open to Remote

DESMOND
ONAM
AI & Data Engineer

Machine Learning Data Engineer with 7+ years of experience building production ML pipelines, deploying LLMs, and architecting cloud-native data solutions on AWS. From fine-tuning transformers to leading cross-continental data science teams.

7+
Years Experience
350+
Engineers Trained
97%
Client Satisfaction
45%
Time-to-Market Reduction
// 01

About

I'm a Machine Learning Data Engineer based in Nairobi, Kenya, specializing in building end-to-end AI systems — from raw data ingestion to production-grade model deployment.

My work spans LLM fine-tuning, MLOps pipeline design, cloud-native ETL/ELT architectures, and real-time data processing. I've deployed systems for government bodies, NGOs, and enterprise clients across Africa, Asia, and the Americas.

Beyond engineering, I am passionate about knowledge transfer — I've trained 350+ data scientists and engineers across Tanzania, Nepal, Bhutan, and Kenya, and co-authored a Master's curriculum adopted by Collège de Paris.

LLMs & GenAI MLOps AWS Certified Data Engineering NLP AI Agents
Python / ML Frameworks95%
AWS / Cloud Architecture90%
LLMs & Fine-tuning88%
Data Pipeline (Spark, Kafka, Airflow)92%
MLOps / CI-CD / Docker / K8s85%
Data Visualization & Analytics87%
// 02

Experience

May 2025 – Present
AI Data Engineer
HavarTech Solutions · Georgia, USA (Remote)
  • Engineered robust AWS data infrastructure using S3, Redshift, Glue, and EMR for large-scale ML training and inference.
  • Built scalable ETL/ELT pipelines with Apache Spark & Databricks — reduced data processing time by 40%.
  • Implemented CI/CD pipelines for MLOps via AWS CodePipeline & Jenkins, improving deployment efficiency by 50%.
  • Fine-tuned RoBERTa-L and Ollama LLMs and integrated them into the product ecosystem with full monitoring.
Dec 2024 – Present
MLOps Data Engineer Consultant / Tutor
Omdena · Bhutan (Remote)
  • Led development of an LLM-powered mental health chatbot adopted by the Bhutanese Government.
  • Designed end-to-end data pipeline for disaster management system (flood, fire, earthquake detection) presented to the United Nations.
  • Trained 350+ professionals across Tanzania, Nepal, and Bhutan with 93% transitioning to live innovation challenges.
Jul 2024 – Nov 2024
Associate ML Data Engineer
Ajua · Nairobi, Kenya
  • Engineered customer experience ML models tracking NPS, CSAT, CES, CLV, and Churn Rate.
  • Built real-time embedded dashboards for clients; migrated the platform to AWS for scalability.
  • Collaborated with Customer Success Engineers to resolve experience challenges and retrain legacy models.
Jun 2023 – Jun 2024
Lead Data Scientist
Nexthikes · Noida, India
  • Led cross-functional teams of data scientists and engineers, achieving 96% client satisfaction.
  • Translated business needs into technical sprints using Agile methodology, increasing revenue.
  • Mentored interns and junior data scientists, earning recognition as Best Mentor.
Jul 2023 – Jul 2024
Data Science Trainer
Henry Harvin · Noida, India
  • Upskilled 215+ students in Python, R, SQL, Apache Spark, and cloud data platforms.
  • Co-architected a Master's in Data Science curriculum now adopted by Collège de Paris.
  • Led 12 real-world data science projects from extraction to model deployment.
// 03

Projects

🧠
Mental Health Chatbot — Bhutan Gov

LLM-powered mental health chatbot fine-tuned on transformer models, with a full data pipeline for continuous retraining. Adopted by the Bhutanese Government to enhance the Gross National Happiness Index.

LLMCrewAINLP AWSFine-tuning
🌊
AI Disaster Management System — UN

Scalable data pipeline ingesting Sentinel-2 satellite imagery and APIs to detect flood, fire, and earthquake events. Deployed on AWS and presented to the United Nations.

Sentinel-2AWS Lambda Computer VisionServerless
📊
Customer Experience 360 Dashboard

Real-time embedded analytics platform tracking NPS, CSAT, CES, CLV, and Churn Rate for enterprise clients. Migrated ETL pipeline to AWS with live reporting to client sites.

Amazon QuickSightRedshift ETLPython
🏔️
COPD Early Detection — Nepal

ML system trained with 120 Nepalese students to identify early causes of Chronic Obstructive Pulmonary Disease in Kathmandu province using climate and health data.

Scikit-learnXGBoost Data AnalysisPublic Health
MLOps Pipeline Optimizer

CI/CD MLOps infrastructure using mlflow, Docker, Kubernetes, and AWS CodePipeline — reducing time-to-market for AI solutions by 45% through automated deployment and monitoring.

mlflowDocker KubernetesAirflowPrefect
🎓
Data Science MSc Curriculum

Co-designed and implemented a Master's in Data Science curriculum covering data engineering, ML, analytics, and cloud computing — now officially adopted by Collège de Paris.

Curriculum DesignLMS Auto-gradingEdTech
// 04

Tech Stack

Languages
  • Python
  • SQL
  • Java
  • JavaScript
  • TypeScript
  • R
ML / AI Frameworks
  • TensorFlow
  • PyTorch
  • Scikit-learn
  • LangChain
  • CrewAI
  • HuggingFace
  • mlflow
Data Engineering
  • Apache Spark
  • Kafka
  • Airflow
  • Prefect
  • Databricks
  • dbt
Cloud & DevOps
  • AWS (S3, EC2, Glue, EMR, Redshift)
  • GCP
  • Docker
  • Kubernetes
  • CI/CD
  • Jenkins
Databases & Storage
  • PostgreSQL
  • Redshift
  • Pinecone (VectorDB)
  • DynamoDB
  • MongoDB
Specializations
  • Large Language Models
  • NLP
  • Generative AI
  • AI Agents
  • Multilingual AI
  • Statistical Modelling
// 05

Let's Build

Open to full-time remote roles, consulting projects, and speaking engagements in AI, Data Engineering, and ML.