Looking for a job

Anslem Akadu

@anslemakadu

Message

AI/ML Engineer at LifeBank building healthcare forecasting systems that reduced stock-out risk by 20%.

Nigeria

Message

At LifeBank (Nigeria) Ltd, I developed forecasting pipelines for blood supply, oxygen, and medical inventory that reduced stock-out risk by 20%. I also created predictive-maintenance models to identify equipment servicing needs earlier.

In LifeBank’s financial systems work, I built hospital finance forecasting models with XGBoost and Prophet, including hybrid forecasts designed for sparse datasets. I later fine-tuned Gemma 1B with QLoRA for hospital finance tasks and migrated an AI system to Qwen, reducing inference latency by 20%.

At 20four7VA, I worked on production-oriented AI systems spanning LLM orchestration, conversational AI, RAG, document intelligence, and workflow automation. My open-source projects include ElevatrAI, a career assistant with skill-to-role recommendations, and a FastAPI and Docker backend for serving GGUF models.

Experience

Work history, roles, and key accomplishments

Current

ML / AI Engineer

Current

LifeBank (Nigeria) Ltd

May 2025 - Present (1 year 5 months)

Fine-tuned Gemma 1B with QLoRA for domain-specific hospital finance tasks, including instruction alignment and domain adaptation. Migrated an AI system from Gemma to Qwen, improving reasoning depth and scalability while reducing inference latency by 20%.

QLoRA Gemma Qwen EC2 BASH GPU Training Model Fine Tuning

Data Scientist

LifeBank (Nigeria) Ltd

Jan 2025 - Apr 2025 (3 months)

Built ML models for hospital finance forecasting across profitability, liquidity, leverage, and efficiency using XGBoost and Prophet. Designed hybrid forecasting models combining 80% naive logic with 20% ML to improve robustness on sparse datasets.

XGBoost Prophet Forecasting Feature Engineering Financial Modeling

AI / ML Engineer

20four7VA

Built and enhanced production-oriented AI systems involving LLM orchestration, conversational AI, RAG, document processing, automation, and business workflow integrations. Developed an AI assistant architecture supporting multiple LLM providers, model routing, and short-term conversational context with Redis.