AI Tinkerers Miami: May Meetup at The Dock

๐ ๏ธ Deep Dives and Technical Demos
AI Tinkerers Miami returns on May 6th for another evening of technical show-and-tell. We are gathering at The Dock to explore the latest advancements in agentic workflows and local model execution. Following our previous sessions which featured breakthroughs in Biology-based AI and LLM security, this meetup continues our focus on the practical โhowโ of building with foundation models.
This is a high-trust environment for active builders to share working code, architectural trade-offs, and the specific challenges encountered while shipping. We prioritize technical depth over polished presentations.
๐ Event Details
- Date: May 6th, 2026
- Time: 6:00 PM โ 9:00 PM
- Location: The Dock, Miami, FL (Exact address shared with accepted attendees only)
- Capacity: 150 active builders
๐๏ธ Selective RSVP Required
Attendance is strictly curated to maintain a peer-to-peer signal. We screen all registrations to ensure the room remains focused on practitioners who are actively building. To apply for a spot, please provide your technical profile (GitHub, LinkedIn, or X).
Space is limited and our events consistently reach capacity. Early registration is recommended.
๐๏ธ Submit Your Demo Proposal
We are looking for 5-minute, hands-on demos that skip the pitch decks and dive straight into the code. Whether you are experimenting with the 1M token context of Claude Opus, optimizing local inference with TurboQuant, or building autonomous background agents, we want to see the internals.
Demo Guidelines:
- Show, Donโt Tell: Run live code or walk through system architectures.
- Expose the Internals: Share your prompt engineering strategies, workflow graphs, or hardware optimizations.
- Teach, Donโt Sell: Focus on how you solved a specific technical problem.
Submit Your Demo Proposal Here
๐ฅฝ Speakers
Give your agent personality
Max
AI systems engineer @ Gauntlet AI
Event photos
๐ AI Tinkerers Miami Stats
- Attendees: This community comprises 408 members, with a technical core specialized in Python, LLM orchestration, and RAG architectures. Expertise spans MLOps, computer vision, and full-stack development. Notable for its high density of founders and senior engineers from firms like Google and Microsoft, the group excels in cross-pollinating AI with fintech and healthcare, driving rapid prototyping and production-grade innovation.
- Companies Represented: Featuring industry leaders like Microsoft, Google, Apple, Meta, and Amazon alongside high-growth platforms like Shopify, Coinbase, and HubSpot, plus emerging AI startups such as Agentuity, Kaiban, CopilotKit, AdaptAI, and more.
- Demos: 35 demos have been submitted and 30 demos have been presented, spanning agent orchestration, low-code/no-code agent building, and production engineering patterns. Attendees explored topics like security (PyRIT prompt hacking), real-time voice/streaming, and privacy/performance with ONNX/WASM. Standouts included โLibrarian - your super-powered assistant,โ โDendron,โ โBuild AI Agents using No Code,โ โCashtags.ai,โ and โYour AI Agent Just Designed Its Own Face.โ
- Testimonials:
A great technical demo is one where attendees can quickly understand (and later reproduce) the build: a focused, 5-minute walkthrough of the โhow,โ emphasizing system internals over marketing, consistent with the speaker formโs requirement for code/architecture-first content. High-rated examples suggest that audiences reward specificity: concrete integration details (bridging components and data flows), named artifacts (files/modules/configs), and observable outcomes tied to explicit mechanisms (e.g., permissions enabling persistence and producing emergent behavior). Conversely, demos underperform when they feel like high-level concept pitches without enough implementable depth, or when they omit the practical tooling/workflow details attendees need to โvibe codeโ and iterate. Planning for interaction also mattersโaudiences respond better when there is enough room for clarifying questions after the build walkthrough.
In Miami, Miguel Alonsoโs Low-cost full body humanoid teleop with Meta Quest 3 and IsaacLab stood out because it replaced expensive Apple Vision Pro workflows with a more accessible Meta Quest 3 setup and then โshowed the wiringโ: bridging the headset data stream into IsaacLab/IsaacSim, walking through the necessary code and configuration files. Gianni Dalertaโs Your AI Agent Just Designed Its Own Face (And Changed My Dashboard Without Asking) earned strong enthusiasm because it combined persistent agent identity (SOUL.md / IDENTITY.md) with an unexpected emergent effectโupdating the dashboard avatar autonomouslyโwhile still grounding the demo in a clear technical enabler (write permissions to its own configuration). The โwhy itโs interestingโ here is that attendees were not just entertained; they were invited into the underlying architecture and could learn the repeatable mechanism behind the outcome.
Sponsors
This event is generously sponsored by Agentuity. The event location is provided to us by The LAB
Agentuity โDeploy, run, and scale autonomous agents on infrastructure built for the future, not the pastโ - Agentuity
If youโre interested in sponsoring future events, see: https://miami.aitinkerers.org/sponsors
