AI-Assisted Product Builder | Business Workflow Automation
I use AI daily to turn operational tasks into working software and reusable workflows. I define requirements, direct AI coding agents, test the results, and iterate. My work combines a business background with hands-on product ownership, document processing, and research automation.
ClearRFP — live product · Public tools · How I work
Creator & Product Owner · Independent, AI-assisted project
ClearRFP is an RFP review and bid-workflow platform: it helps organize requirements, surface risks and compliance items, and coordinate the work needed to prepare a response.
- Took the initial product from idea to production in approximately six weeks using AI coding agents; owned product scope, requirements, and user-experience decisions.
- Secured an initial business pilot for a 7–10-user group to gather product and usability feedback.
- Directed development of document analysis, bid/no-bid review, and Word and Excel outputs.
- Shaped workflows around task ownership, review steps, proposal preparation, and submission evidence. People retain approval and decision-making authority.
The application source is private. The public website provides a product overview; the pilot is an initial feedback milestone, not a claim of paid adoption or measured business impact.
Document preparation for AI. Created a PDF-to-structured-text workflow with local extraction, page-level quality checks, and selective cloud text recognition. This is a separate, private project.
Customer discovery — AddinFab. In an academic startup team, conducted six customer-discovery interviews and surveyed 92 manufacturers. Contributed research, functional requirements, and positioning for a concept covering 3D-printing orders and production management.
Website and inquiry intake — Accirva. Delivered the corporate website through AI-assisted development, including structured service information and a consultation-request form connected to email delivery. My contribution was the website project.
These are focused integrations built with AI coding agents around existing services and libraries. They support repeatable research, document, and browser tasks; they are supporting tools rather than the main product portfolio.
| Project | Practical purpose |
|---|---|
| notebooklm-cdp | Connect an authorized Chrome session to NotebookLM workflows; add convenient note access and Markdown export on top of notebooklm-py. |
| exa-search-cli | Make Exa search, page-text retrieval, and research tasks accessible from scripts and AI-agent workflows. |
| browser-agent-cli | Launch a visible Chrome Beta session with a separate agent profile for supervised browser tasks on macOS. |
| yandex-search-cli | Bring Yandex search, cited answers, and related research services into structured command-line workflows. |
Each repository documents its scope, dependencies, and setup. Provider services supply the underlying search, browser, or AI capabilities.
Understand the task → define requirements → direct AI-assisted implementation → check the result → iterate.
I work with tools such as ChatGPT, Claude Code, Codex, and NotebookLM. My contribution is framing the problem, choosing the workflow, specifying behavior, guiding implementation, and checking usability and outputs. Coding agents are part of the implementation process, not a substitute for human judgment.
For document and research workflows, I keep source checking, review, and final business decisions explicit. I focus on practical use cases rather than recommending another tool before understanding the work.
Nolan Vale Tools is the label for my independent public projects.