Current Projects & Future Ideas

ClearLease started as a lease-abstraction pipeline: a lease agreement or master lease document in, a structured and source-cited abstract out, so every extracted value can be traced back to the exact page and clause it came from instead of taken on faith. The engine itself is done and has already been through a full facelift, a synced source viewer (click a citation, see the exact clause highlighted on the original page), and two real performance wins (smart OCR routing that skips Azure entirely on clean digital PDFs, and batched concurrent extraction that cut a 260-field run from 260 individual API calls to roughly 44). The showcase build is a custom Next.js app, demoed through screenshots and a recorded walkthrough rather than a public link, on purpose, since a public link would route strangers' traffic through my own paid API keys before there's any billing or accounts to gate it.

What's actually next

• Extra security and data-handling hardening: password hashing, session strategy, secrets management, data retention policy, and multi-tenant isolation confirmed at the database level, not just the UI.
• Real business formation, on paper, before it ever takes a dollar: an LLC, an EIN, a dedicated business bank account, a registered domain, and drafted Terms of Service and a Privacy Policy.
• Accounts and cross-device sync (log in at home, pick up at the office), a capped free trial instead of an open-ended free tier, and Stripe billing priced against the actual measured per-document API cost rather than a guess.

Two ways this could actually ship

Model 1 is a SaaS I run myself: my own API keys power everything, customers pay a subscription, and their lease data flows through my backend. Model 2 is a source license: a firm buys the source and runs it entirely in-house on their own Azure and Anthropic keys, so their data never touches my systems at all. There's also a LOTR-themed edition of ClearLease, an internal-only showcase version I can present to a company that wants to see this exact system implemented and themed to their own brand rather than run through my website at all. That edition doubles as the pitch for the source-license model, literally: hire me, and I'll implant this exact system inside your company, fully internalized and confidential, no public website required. Check the video player under Projects > ClearLease, the walkthrough video has a second part where I go through the LOTR-themed edition in detail (recording and uploading soon).

The bigger idea, parked for now

ClearLease is the starting point for a broader idea I've been sitting on: expanding this into an actual company called ClearRE, which would become my main LLC, domain, and company home base. Under that umbrella, ClearLease handles lease abstraction, ClearWriting would be an AI-assisted underwriting reviewer, ClearAgent would source deals off CoStar, Zillow, and similar platforms, ClearCust would handle CRM management, and there are a handful of other ideas beyond those. Each one would work the same two ways ClearLease does: a company can hire me to build and implement the same tool internally under contract, fully theirs, or they can use it as a web-based service I host and run on my own APIs, and they simply pay for it. This is a future idea, not something I'm building yet: I'm waiting to see whether ClearLease itself gets real traction once I finish and launch it before I spend money legitimizing and building out the rest of these tools.

In a nutshell, I'm training trading agents targeting a minimum 80% win rate across futures, options, and other instruments. Still in the concept stage and staying that way here, not posting process, mechanics, or strategy publicly. If this is something you'd want to talk about, get in touch.

I've been brought on as an outside consultant on an early-stage sports and fitness training concept. The core idea is a standardized testing protocol that measures an individual athlete across a handful of physical categories, speed, agility, power, strength, and endurance, and turns that into a single, trackable score with more detailed sub-scores underneath. The philosophy driving it is simple and repeatable: test, identify weaknesses, train around them, retest on a set cadence, and track the trend over time rather than guessing whether training is actually working.

My role has been advising on the web and product side, thinking through what a real front-end and back-end build looks like versus a basic marketing site, how a future mobile app might get wireframed and prototyped, and how to think about accounts, testing history, and data over time if this grows past a single trainer's notebook. Details here stay intentionally general out of respect for the people building it.