Partnerships
AI Architect — Gemini + ADK enterprise AI CoE practice
Gemini already in daily multimodal delivery. Google ADK + A2A protocol already in the workshop enterprise lane. The conversation is Cloud Partner Advantage entry + workshop credits.
A deeper conversation is open with Google. The proposal page comes online when both sides are ready to publish.
Context
Gemini is part of my daily multimodal delivery — text (Gemini 3 Pro for long-context grounding), image (Nano Banana 2 / Imagen via Higgsfield MCP), and video (Veo) all sit inside the working set alongside Claude and the OpenAI stack. Antigravity is the multi-model agentic dev surface where Gemini 3 Pro and Claude Sonnet refactor side-by-side — multi-model parallel comparison as daily practice, not benchmark theater. The Google Agent Development Kit and the A2A protocol are not a slide — they are the enterprise lane of the Build First AI Agent workshop and the explicit enterprise lane of the Starlight Agent Lab three-lane portfolio (Vercel AI SDK / Claude Agent SDK / Google ADK).
The strategic alignment here is about a Google-native AI CoE practice with workshop and curriculum distribution. Pursuing Google Cloud Partner Advantage status. The currency that matters more than commission is attendee credits for the workshop cohort and reference-architecture surfaces with Deploy-to-GCP CTAs. FrankX already has public builder channels where the topic fits naturally.
Working reality
Verifiable today. The proposal section below builds from here, not in place of it.
Gemini in daily multimodal delivery
Gemini 3 Pro, Nano Banana 2, Veo, and Imagen sit inside the working set used for daily multimodal output — research, image (via Higgsfield MCP), video, and long-context code. Not a benchmark exercise. The daily-use stack.
Antigravity for multi-model agentic dev
Antigravity is where I run multi-model parallel refactors — Gemini 3 Pro and Claude Sonnet side-by-side on the same task, pick the best, merge or compose. Multi-model comparison as daily practice. Native browser pane and Vercel deploy integration round it out.
ADK + A2A in the Build First AI Agent workshop
The Google branch — Agent Development Kit and the Agent-to-Agent protocol — is one of the enterprise tracks taught in the workshop. Cohort gets hands-on ADK and Agent Cards, not framework theater.
/workshops/build-first-ai-agentPublic A2A protocol guide
Agent Card / A2A protocol specification documented end-to-end on /guides/agent-card-a2a-spec — written from hands-on use, with the wire-level detail an enterprise architect needs.
/guides/agent-card-a2a-specStarlight Agent Lab enterprise lane = ADK
The three-lane portfolio explicitly positions Google ADK as the enterprise lane alongside Vercel AI SDK (web) and Claude Agent SDK (reasoning). Multi-cloud by design, ADK by enterprise default.
Google Cloud Partner Advantage — pursuit underway
Pursuing Google Cloud Partner Advantage status. The attendee-credit channel for the workshop cohort is the higher-value mechanic — credits convert to durable Gemini and Vertex usage across the EMEA architect funnel.
Conversation open
The proposal page comes online when both sides are ready to publish. Until then, the working partnerships hub explains how I collaborate at this depth.
If you're reading this and the conversation is yours to open on the Google side, the calendar link is below.
Last updated