Since ChatGPT launched in 2022, I have tracked AI developments closely and driven practical adoption across my organization. Below are the major AI initiatives I spearheaded.
Contract and document reviews require verifying signatures, clauses, and conditional logic. I collaborated with team leads to craft detailed, reusable prompts for Microsoft Copilot. These prompts now extract and validate key information in minutes. A review that once took 15–20 minutes now completes in 2–3 minutes. This cut processing time dramatically and boosted throughput without quality loss.
Our Google Sites knowledge bank required manual searching. I proposed and built an AI-powered internal bot using Chatbase, fed with all SOPs, process documents, and custom prompts. Teams in non-voice processing and customer service now ask natural-language questions daily and receive instant, accurate answers. I maintain the bot: ingest updates, refine prompts, add sources, and test new models (OpenAI, Anthropic, Grok, Gemini). Over two years, answer accuracy has soared, making the bot a reliable virtual SME for the entire team.
Traditional SOPs and client portal guides took weeks or days. I created a fast dictation-to-document workflow: record voice notes (15–30 minutes for long processes), transcribe via phone AI, feed transcript to Microsoft Copilot with a structured prompt to clean, organize, format as SOP/guide, and insert screenshot placeholders. Final screenshot insertion takes 15–20 minutes. Full documents now emerge in hours instead of weeks, revolutionizing content creation speed and quality.
As product manager/scrum master, I gather requirements and write Jira user stories. I record every requirements meeting, transcribe it, and prompt Copilot (or Amazon Q) to extract complete requirements and generate polished user stories directly in Jira. Manual note-taking, synthesis, and story writing—previously hours or days—now finish in minutes after the call ends. Nothing gets missed because the full transcript drives the output.
Creating and revising process flowcharts in Visio used to take hours plus rework time. I shifted to Amazon Q: feed it the meeting transcript or step description, and it produces clean, professional flowcharts in minutes. Tweaks require only new prompts—no dragging shapes. Accuracy improved because the diagram reflects the exact discussion transcript. This delivers faster, more precise visuals for developers and stakeholders.
For client-facing portal redesign, I used Anthropic’s Claude to generate multiple layout and UI iterations by describing desired elements and user flows. These AI-generated concepts formed the starting point. Our UI/UX and development teams refined them into the final design. This accelerated ideation and brought fresh, modern ideas forward quickly.
These initiatives demonstrate measurable gains in speed, accuracy, cost, and team productivity through targeted AI deployment.
Home ⬩ Work Experience ⬩ Initiatives ⬩ AI Initiatives ⬩ Education ⬩ Contact