CPA | AI Practitioner
As a Certified Public Accountant and AI Practitioner, I bridge the crucial gap between accounting and finance and transformative artificial intelligence. My mission is to empower businesses by optimizing accounting and operational workflows with advanced automation, unlocking new levels of efficiency and insight.
Building expertise across accounting, finance, education, and innovative technology adoption
Licensed CPA with a proven track record in financial reporting, auditing, and tax strategy across diverse industries. Successfully led teams to implement best practices and ensure robust regulatory compliance.
Passionate former accounting educator dedicated to demystifying complex financial concepts and fostering professional growth. Mentored numerous aspiring CPAs, guiding them to certification success.
Championed digital transformation within traditional accounting environments, spearheading initiatives that integrated automation and modernized workflows, resulting in substantial efficiency improvements.
Developing practical AI solutions to transform accounting, finance, and tax operations and create tangible business value
Self-hosted various LLMs and explored fine-tuning them on M3 Mac hardware. The development of a fully realized application presented unique challenges; however, the process provided invaluable insights into the intricacies of gradient checkpointing, memory optimization, and ensuring model stability. Gained significant practical knowledge regarding hardware constraints, the critical role of dataset quality, and the nuances of robust fine-tuning approaches for achieving factual accuracy.
Built a tax assistant using Retrieval Augmented Generation with quantized Llama 3 8B. Leveraged LangChain, Sentence-Transformers, and FAISS-CPU to create an efficient question-answering system. Gained deep insights into context management and vector database optimization.
Created six Python scripts with GPT-4 assistance that reduced 5+ hours of manual accounting work to under 30 minutes. These automation solutions demonstrate the transformative power of AI in traditional workflows, achieving dramatic efficiency improvements while maintaining accuracy and compliance.
Deep-dived into transformer model architectures through hands-on visualization and analysis. Explored tokenization, embeddings, self-attention mechanisms, and feed-forward networks to understand how modern language models process and understand text. This foundation enables more informed AI solution design and implementation.
Successfully developed and completed an automated workflow for sending BIR DAT file attachments (SLSP, SAWT, and QAP) using n8n, significantly streamlining the submission process. The system extracts client data from spreadsheets, standardizes TIN formats, groups clients with available DAT files, and dynamically generates personalized email subjects. It then retrieves and consolidates the required files, dispatches them via email to the BIR eSubmission portal, and includes built-in validation and detailed reporting for transparency and error tracking. This robust automation, developed with the assistance of Gemini 2.5 Pro and Claude Sonnet 4, made a complex task more efficient and reliable.
Created a course titled “Introduction to Generative AI with Applications,” designed specifically for accounting and business students to explore the fundamentals of generative AI and its real-world uses. Interactive learning materials were developed to ensure the course is both engaging and practically impactful for students and the university alike.
Ready to explore AI's potential in finance or collaborate on innovative projects? I'd love to hear from you.