Wild Eye AI
Intelligence engines for complex, consequential decisions. The work turns ambiguous information into evidence operators can act on, whether the domain is finance, law, or anywhere guesswork is too expensive.
AI-native systems / consequential work
Signal
to system.
Frontier AI built with care and applied with discipline, for decisions where guesswork is too expensive.
Built with judgment.
Designed for scrutiny.
SCIENCE / TECHNOLOGY / LAW
01 / Thesis
Business understanding drives the build. Domain expertise fortifies it.
I translate business needs and goals into functional, valuable, secure software. Keeping domain expertise close to the work sharpens the requirements, exposes assumptions, and helps define both a good result and what the system must protect.
Read the thinking behind the work02 / Work
Systems built where being wrong is expensive.
Fortress Intel
A proxy campaign platform built for the work solicitor firms run. It covers the full lifecycle: pre-campaign analysis, shareholder intelligence, outreach, real-time risk management, document production, and post-campaign compliance archival. Frontier AI is embedded throughout to lift the quality, speed, and consistency of deliverables. The goal is to help experienced teams secure participation more efficiently, with improvements in cost, timing, and adjournments measured in real campaigns.
Read The Participation Gap on outreach and shareholder intentFortress Evidence Intelligence Platform
An AI-native system for legal and expert-witness workflows that grounds conclusions in verifiable evidence, provenance, and human review. Fortress is being developed to make the path from source material to consequential assertion inspectable.
See how evidence integrity informs our Legal AI Assurance workClickPoint Software
President and COO of a lead management platform, originally incubated within Steinway and spun out in 2013.
Steinway & Sons
Former CTO. Founded the Music Technology Division and led early R&D for the company's first self-playing piano system.
Legal AI Assurance / California law firms
Use AI with controls you can defend.
Wild Eye AI helps law firms turn AI obligations into working controls for research, drafting, evidence review, court filings, and agentic workflows. The work connects policy, technology, supervision, and proof so responsible practice is visible in daily operations.
Security / built in from the beginning
Trust has to be
built into the system.
Security is integral to how we design, evaluate, and improve AI-native systems. Controlled access, human oversight, and verifiable remediation are part of the work.
Our defensive security work is supported by Trusted Access to restricted frontier AI security capabilities, applied within documented authorization and accountable human oversight.
- Explicit authorization01
- Traceable evidence02
- Human oversight03
- Verified remediation04
03 / Writing & open source
Craft you can inspect.
Results you can verify.
Select essays, methods, and best practices, published openly under Creative Commons and MIT so others can build on them. The throughline matches the work itself: craft you can inspect, and results you can verify.
Featured essay / updated October 2026
The Moat Was Never the Code
Translating business needs into valuable software takes judgment. Domain expertise fortifies the work and helps define what the system must protect.
Research essay / Fund governance
The Participation Gap
How better outreach and Advanced Voting Instructions could help teams reach shareholders, preserve their stated choices, and make participation easier.
14 min read 19 referencesLabs / Open resource
The AI Builder Playbook
Let AI build your prototype. Don’t let your prototype own you. Standards, prompts, and runbooks for keeping control of an AI-built application.
Version 0.1.1 CC BY 4.0 / MITEssay / in progress
Further observations
A second piece on judgment, domain knowledge, and building intelligence systems.
In progress04 / Background
Henry V. Alfano
Founder and CEO
A technology executive and practitioner with over twenty-five years building web applications and SaaS. His approach is shaped by work in scientific research, software development, intellectual property, and dispute resolution. He holds a master's degree and a graduate certificate from the USC Gould School of Law.
The legal training informs how he builds: systems designed from the start to hold up under scrutiny, scale without exposure, and create durable advantage rather than short-term lift.
05 / Contact
Start a conversation.
henry@wildeye.coFor security concerns, use the Security Contact form.