Sikirulahi Abdulkareem

AI/ML Engineer and Python Developer with 4+ years building production AI applications and robust backend systems. Architect end-to-end solutions from custom LLM integrations and multi-agent workflows to RESTful APIs. Delivered measurable impact including 80% workload reduction through AI phone agents, 70% accuracy improvements via hybrid RAG architectures, and 25% cost savings through model optimization.

Expertise spans LangChain/LangGraph, PyTorch/TensorFlow, FastAPI/Flask, vector databases (Pinecone), and CI/CD pipelines.

AI/ML Engineer at Optimus AI Labs (2025-01 – Present)

• Designed and deployed an AI-powered dental phone agent using Twilio, enabling 24/7 patient service. The system autonomously handles appointments and inquiries, reducing front desk workload by approximately 80%
• Architected a hybrid RAG knowledge base system using PostgreSQL and Pinecone and implemented advanced re-ranking techniques, boosting response accuracy by 70% and reducing query latency by 60%
• Engineered reliable multi-agent AI workflows with LangGraph/LangChain, incorporating validation layers and feedback loops to ensure robust performance in production
• Managed the full LLM lifecycle, from fine-tuning models with PyTorch/TensorFlow to production deployment and implementing token optimization strategies that reduced inference costs by an average of 25%
• Conducted comparative analysis of major LLM providers (Claude, GPT, Gemini, etc.), benchmarking performance to optimally align model capabilities with use-case requirements for accuracy, speed, and cost
• Researched and validated multiple RAG architectures (e.g., hybrid, multi-query, recursive retrieval) to determine the optimal strategy for specific use cases, improving overall system relevance by up to 40%
• Developed scalable backend APIs with FastAPI/Flask and MySQL/Postgres to support AI services and data operations, ensuring high availability and integration
• Automated data pipelines for web scraping and preprocessing with Pandas and Python, transforming unstructured data into curated datasets for model training
• Deployed and optimized AI solutions on Railway and Runpod within Linux environments, ensuring scalable and reliable cloud operations

AI Engineer/Python Developer at Syncline Tech Labs (2023-08 – 2024-11)

• Built custom AI applications for clients by integrating GPT-4, Claude, and PaLM, everything from document processing tools to code analysis assistants and interactive Q&A systems
• Created automated agent workflows with LangChain and LangGraph that handled complex multi-step tasks, like generating reports and building intelligent code refactoring tools
• Developed RAG systems for technical documentation, implementing semantic search and context retrieval to help clients get accurate answers from their knowledge bases
• Built backend APIs using FastAPI and Postgres to power client applications, making sure everything integrated smoothly with their existing systems
• Set up CI/CD pipelines with GitHub Actions to speed up deployment and allow for quick iterations on client projects
• Cleaned and prepared text data using NLP techniques; tokenization, entity extraction, semantic clustering to get datasets ready for model training and fine-tuning
• Debugged client code and optimized performance, finding bottlenecks and implementing fixes that made their systems faster and more efficient
• Wrote clear technical documentation for every project handover so clients could maintain and extend the solutions on their own

Data Scientist at Hamoye AI Labs (2023-05 – 2023-08)

• Built and deployed a customer churn prediction model using Python, scikit-learn, and pandas, achieving 87% accuracy with real-time Streamlit interface to access the churn probability scores
• Built economic indicator prediction models for Nigerian remittance analysis using time series analysis and regression techniques, forecasting GDP growth, remittance-to-GDP ratios, and unemployment rates with Streamlit deployment
• Analyzed Nigerian economic datasets using pandas and NumPy to identify correlations between remittance flows and macroeconomic indicators for predictive modeling

B.Eng in Mechatronics Engineering – Federal University of Agriculture, Abeokuta (2018-10 – 2024-09)