Ayomide Abayomi-Alli

– Led end-to-end system design and feasibility assessment: Translated organizational needs into a detailed strategic roadmap that addressed user adoption, governance, and accountability requirements, and delivered a phased deployment plan within a Microsoft Azure tenant. – Designed and deployed a Slack-based chatbot to automate donor proposal drafting: Automated data collection, multi-agent AI summarization, and human-in-the-loop collaborative drafting to accelerate donor proposal preparation (targeting 60% time reduction) – Key technologies: N8N (workflow automation) | Slack API | OpenAI API | Google API || Web & document extraction

– Curated and formalized large-scale geospatial datasets: Defined three core spatial concepts (direction, topology, distance) to create a structured benchmark for probing LLM geographic commonsense. – Designed and implemented an automated data pipeline: Utilized SPARQL, GraphDB, and Python to extract almost a million rows of data from RDF triples into balanced atomic and composite datasets for model probing, in a Linux server. – Benchmarked and analyzed state-of-the-art LLMs: Utilized zero- and few-shot prompting to uncover weaknesses and provide actionable insights for integrating LLMs with explicit geospatial reasoning systems. – Key technologies: YAGO2geo, GraphDB, OCCIDATA, LLAMA, Mistral, GeoSPARQL

– Led the development of LLM chatbot: Drove the end-to-end development of chatbot using Azure Function tools, OpenAI, and Flask API to improve access to real-time insights, contributing to the 5.5 million ATC growth. – Implemented churn mitigation model: Boosted liquidity, customer retention, and reactivated over 20,000 corporate accounts, contributing to a NGN 2.29 trillion increase in customer deposits. – Led user training and technical adoption: Trained 500+ banking staff across multiple branches, achieving 75% user adoption of new BI solutions. – Deployment and maintenance of the 'Wemawiki' AI chatbot: Periodically update and deploy data to the Azure container in production. Reducing operational turnaround time for over 2,000 customer-facing staff members to foster a more efficient work environment. – Built customer segmentation clustering model: Drove a 1,428% increase in loans, and a 2,150% rise in female SME transactions, contributing to a NGN 1 trillion loan growth in Q3, 61% of the total interest income in Q2. – Key technologies: Azure OpenAI | SSMS | Azure ML | MongoDB | Advanced analytics | Postman | Git | Basic MLOps

– Developed customer churn classification model and monitoring dashboard: Partnered with the branding and marketing team, and delivered a churn project to track churn trends and increase customer retention by 31.33%. – Proposed KPI-driven growth strategy for the savings platform: Delivered an innovative, data-backed proposal to executive management, resulting in over 800 new stash accounts in 2023. – Designed a customer feedback and sentiment dashboard: Scraped data from social media and CRM systems for sentiment analysis to identify recurring user pain points, increasing customer satisfaction scores by 48% in six months. – Revamped existing ELT data pipelines: Formulated comprehensive data validation rules, ensuring consistent data flow to Power BI reports; improved data quality scores by 30% within one quarter. – Developed gamification algorithm: Segmented customers using ML, and cohort analysis to track retention, boosting conversion rates by 20% and retention by 17% in 6 months. – Implemented a rule-based fraud detection framework: Using graph databases and link analysis to identify connected risk patterns, improving operational efficiency, and preventing financial losses from fraud. – Key technologies: Python | MySQL | PostgreSQL | Google Bigquery | Power BI | Google Analytics | Freshdesk | ERP

– Developed a financial monitoring dashboard: To track income flow and ensure efficient financial controls, leading to a 16.8% increase in profit before taxes for the first half of 2022. – Built and managed Tableau BI reports: To provide non-technical stakeholders with immediate access to necessary data, reducing manual reporting time by 15 hours per week. – Key technologies: Tableau | Microsoft Excel | PowerPoint | MySQL | Microsoft 365