Synthograsizer
About

The person behind the instrument.

Built for creative people working with creative agents.

"The medium is the message."

— Marshall McLuhan, Understanding Media: The Extensions of Man, 1964

§ Now
Currently Building

Synthograsizer Suite

A solo-designed, open-source browser-based AI workflow platform. Agent Studio (multi-agent LLM orchestration with per-turn trace viewer), a portable JSON template DSL spanning prompt assembly + batch evaluation + agent definitions, and real-time integrations across MIDI, OSC, and WebRTC. Recently selected for the Daydream Scope Interactive AI Video Program (Cohort 3).

7
Studio Tools
50+
JSON Templates
46+
Agent Personas
Cohort 3
Daydream Scope
Launch the suite View on GitHub

A creative toolkit, authored by hand.

Synthograsizer is an interactive web application for structured prompt engineering and real-time generative art control — inspired by the tactile experience of hardware synthesizers. It features a D-pad navigation model, MIDI and OSC support, batch generation, and embedded p5.js sketches for live visual experimentation.

The suite spans seven tools — a prompt sequencer, glitch studio, multi-agent chat room, browser DAW, metadata manager, and more — all sharing a common JSON template format so ideas can move freely between them.

About the Creator

Alexander Heemskerk is a creative technologist and AI workflow architect passionate about making advanced technology accessible, intuitive, and inspiring. His background spans generative art, interface design, and creative AI — with a focus on building tools that empower artists, designers, and curious minds to explore new forms of expression.

Recent work focuses on multi-agent orchestration with observability primitives — turn-order routing, per-turn trace capture, and consensus detection across LLM agents — built on a portable JSON template format that doubles as an evaluation harness for batch prompt runs.

Alexander believes technology should feel playful and empowering — not intimidating. His work bridges the gap between cutting-edge AI and creative practitioners, centered on human-centered design and sustainable creative ecosystems.

Core Skills & Interests

Engineering
AI Workflow Design
Generative AI Systems
Node-based Architecture
Creative Coding
UX for Complex Systems
Interactive Media
Practice
Product Strategy
User Research
Technical Writing
Community Building
Blockchain & Attribution
Creative Problem Solving

Selected Experience

Browser-based AI workflow platform: multi-agent orchestration (46+ personas, autonomous turn-taking, trace viewer), JSON template DSL spanning prompt assembly + batch evaluation + agent definitions, real-time integrations across MIDI, OSC, and WebRTC.
Emerge — Aug 2025 – Apr 2026. Designed the image and video generation workflows powering @tryemerge, a Farcaster-native AI agent that turned a user's profile photo plus a natural-language prompt into personalized social-media content. Engineered and iteratively refined the agent's LLM system prompts across four production revisions to handle action-routing (create / like / recast), transaction-safety guardrails, and conversation-length edge cases — each pass informed by observed production failures. Authored auxiliary prompts for workflow auto-naming (structured JSON output) and token-branding extraction. Produced launch assets including a flagship influencer-collab workflow and a 700+ asset batch generated from real user prompts.
Platform impact (team)
30K
Total users
70K+
Pieces generated
#1
Farcaster · 5 wks
$90K+
Revenue
Collaboration with Michael Beets · May 2024. Built the generative pipeline for Here and Now Edition 6, the final installment in Beets' immersive Tezos blockchain series. User interactions inside a Unity-based 3D gallery drove trait selection, which fed a p5.js sketch, a detailed generative prompt, and personalized LoRAs — producing a unique image of each participant emerging from Plato's Cave. Powered by the EmProps OpenStudio API in real-time.
EmProps.ai · Mar 2023 – Aug 2025. Designed and optimized modular AI workflows integrating Stable Diffusion, SDXL, and LLMs into accessible creative tools used by hundreds of creators. Identified friction points in complex AI interfaces and built solutions to make advanced capabilities usable by non-technical users. Contributed to product roadmap and feature prioritization; conceptualized frameworks for content attribution, collaborative creation, and adaptive workflows. Led the technical transition from Legacy to EmProps 2.0. Primary technical resource for the creator community on prompt engineering and JavaScript/p5.js creative coding.
Collaboration with ClownVamp · Sep 2022 – Mar 2023. Designed and built the backend infrastructure for a blockchain-powered choose-your-own-adventure narrative on Tezos, enabling branching story choices to be resolved and recorded per-wallet. ClownVamp is a conceptual AI artist whose work has been exhibited in New York, Paris, and Los Angeles, and featured in Artnet, Outland, and Surface Magazine.
Ottawa International Animation Festival / IFFO · Jul 2020 – Oct 2021. Three consecutive roles spanning the full pandemic pivot to virtual programming. Orchestrated the digital transformation of major international festivals — selected, deployed, and operated webinar platforms, A/V streaming software, and interactive networking spaces for a global audience. Served as primary technical point-of-contact during live events, troubleshooting in real-time for presenters from major international animation studios. Developed onboarding documentation and training materials; analyzed engagement metrics to optimize future programming.

Work With Me

Open to collaborations, creative technology consulting, and conversations about the future of AI and creativity.

Get in touch

Browse the Art

Generative art, NFT drops, and creative experiments across platforms — all linked in one place.

View gallery