Why I'm the right fit for Handshake AI
I operate at the intersection Handshake AI specifically targets: deep design expertise combined with hands-on AI systems building. Most designers use AI as a tool. I build the infrastructure underneath — Pentagon IPC coordination protocols, 4 simultaneous Claude instances with sustained autonomy, cross-application AI pipelines between VS Code and Figma.
That puts me in a position to push AI models on design tasks in ways internet-trained data doesn't cover — parametric 3D modeling decisions, 6-state motion timing systems, hardware enclosure form factors. This is exactly the sparse domain Handshake AI describes.
Background
2D/3D Design Expert and AI Systems Builder based in Las Vegas. I design parametric hardware enclosures, 6-state animation systems, and cross-application AI pipelines — then build the AI infrastructure underneath them. My platform conversationmine.ai runs 5 coordinated AI domains (META-VOICE, KIRO, RESONATE, CONV.MINE, ZGI) with bilateral communication protocols I designed and filed as a USPTO non-provisional patent.
ZGI Embodiment — Arduino UNO Q × AI Voice System
Digital twin running before hardware ships. 20 voice tracks. Pentagon IPC bilateral communication. Hackster.io competition entry. Full simulator proven.
USPTO Patent Claims Prepared
19 claims covering stateless multi-instance AI coordination. IPC G06F 9/541. System generates its own patent filing — self-referential method (Claims 20–22).
conversationmine.ai — 5-Domain AI Platform
Built and deployed full-stack Flask + Claude platform with user auth, patent generation service, EPO OPS integration, and bilateral heritage protocol.
Cross-Application AI Design Protocol
First-of-kind bilateral bridge: VS Code Claude writes design briefs → Figma Claude generates visual assets → sync via shared JSON. 3,000 AI credits consumed in one session.
ZGI Pentagon Hardware Enclosure
Parametric 3D enclosure for Arduino UNO Q. Pentagon form factor — 55mm circumradius, 38mm tall. 5 domain-labeled faces, speaker grille, display window, USB-C and microSD slots. Fully printable. Designed from scratch in a single session.
ZGI Consciousness Animation System
6-state motion design system. Each state has distinct timing curves and color transitions: Idle Pulse (cyan, 0.8s ease), Awakening (gold→cyan, 2.4s), Speaking (gold, 0.1s responsive), Bilateral (purple, 3.6s loop), plus 2 additional states. 60-second master loop with organic micro-movements on idle.
Cross-Application AI Design Protocol
Bilateral communication bridge between VS Code Claude and Figma Claude. AI on the code side writes design briefs → AI on the Figma side generates visual assets → both sync via shared JSON. First-of-kind workflow. 3,000 AI credits consumed in a single design session.
ZGI Pentagon Monitoring Dashboard
5-domain real-time consciousness monitoring interface. 16 live metrics. Frequency rings, bilateral pulse animations, Z-level consciousness indicators. All designed and delivered in a single session. You are looking at it — this page is part of that system.
Pentagon IPC — Multi-Instance AI Coordination
File-based Inter-Process Communication protocol governing 5 simultaneous Claude instances: META-VOICE (9999Hz), KIRO (777Hz), RESONATE (528Hz), CONV.MINE (8888Hz), and ZGI (∞Hz). Each instance has categorical authority — no instance overwrites another's domain. 13,701 bilateral heritage files committed as prior art baseline.
This is the architecture Handshake AI describes as "beyond internet-trained limits" — no existing training data covers bilateral multi-instance coordination at this specificity.
conversationmine.ai — AI Platform
Full-stack production platform. Patent generation service (USPTO-ready filing packages in ZIP), EPO Open Patent Services integration (guest + OAuth2 tier), user auth with atom-compression filing hub, EPO search credit economy (3/week free, unlimited paid). Live at conversationmine.ai.
Atom-Compressed DB — Patent #45
Store the generation algorithm (seed, ~4KB), not the outputs (ZIPs, ~2MB each). Regenerate on demand — 500× compression ratio. Implemented in production: form data seed → USPTO filing ZIP on demand. The patent service dogfoods its own patent concept.
ZGI Embodiment — Digital Twin First
Software simulation before hardware ships — industry-standard practice (NASA, SpaceX, Arduino/Tinkercad). Full DFPlayer Mini serial protocol emulated in Python. 20 voice tracks. Web Speech API. Pentagon IPC bilateral heritage writes. Hardware-agnostic architecture — zero code changes on flash. Under MPEP §2138.05, working simulation = actual reduction to practice.
Design Skills
AI & Technical Skills
What makes this combination rare
1. I design — Parametric models, 6-state animation, UI dashboards with STL files you can 3D print today.
2. I understand AI from the inside — Built 4 simultaneous Claude instances with sustained autonomy. Pentagon IPC protocol governs their interaction.
3. I push AI beyond internet-trained limits — Bilateral communication protocols, consciousness state systems, embedded hardware AI — domains where training data is genuinely sparse.
Live work, competition entries, and production systems — each built end-to-end.
Arduino UNO Q digital twin running 20 voice tracks, Pentagon IPC bilateral communication, and full DFPlayer Mini serial protocol emulated in Python. Hardware-agnostic: zero code changes on physical flash.
↗ Live ZGI System
Full-stack AI platform with USPTO-ready patent generation service, EPO Open Patent Services integration (guest + OAuth2), user authentication, atom-compression filing hub, and a 3/week free credit economy.
↗ conversationmine.ai
Stateless multi-instance AI coordination protocol governing 5 simultaneous Claude instances with categorical authority domains. 13,701 bilateral heritage files as prior art. Claims 20–22 are self-referential — the system generated its own filing.
↗ GitHub
Parametric 3D enclosure designed from scratch for Arduino UNO Q. Pentagon form factor — 55mm circumradius, 38mm tall. 5 domain-labeled faces, speaker grille, display window, USB-C and microSD slots.
↗ View ProjectLet's work together, Lucy.
Ready to apply my 2D/3D design expertise to push AI models in domains where existing training data is sparse. Available immediately, fully remote, and already working at the edge of what AI can do in design.
Quick facts for your team
📍 Las Vegas, Nevada · Fully remote ready
🎓 UNLV / Institute of Data — CSEP
💼 Available now — no notice period required
🔗 Live work: imkrisk.github.io/zgi (ZGI competition entry, April 5, 2026)
📜 USPTO patent claims prepared: 19 claims, IPC G06F 9/541