About Omni Aura

We build the memory layer for AI that stays with you. The story started five years before any of it shipped.

#The team

Two founders. Four builders across mobile, operations, creative, and sales. The same team stayed through the pivot from hardware to software.

Portrait of Peyton Spencer

Founder

Peyton Spencer

Business, operations, and engineering.

Bio

Leads business strategy, operations, and fundraising while co-leading engineering. Came up through music tech, building synthesizers and a real-time light engine before pivoting to AI systems. Ships the Go backend, agent pipelines, and SolidJS frontend that power HeyDitto.

Portrait of Omar Barazanji

Founder

Omar Barazanji

AI systems and technical architecture.

Bio

Leads AI and technical architecture. Was studying transformers and GAN image generators long before ChatGPT. In 2023, built the full Ditto smart-home stack on a Raspberry Pi, including an Image RAG vision server that went live eight days before OpenAI released GPT-4V. Designs the memory graph and agent orchestration layer.

Portrait of Romanee Spencer

Romanee Spencer

Chief Operating Officer

Bio

Keeps operations and strategy aligned with what users want. Background in Accounting, MIS, and real estate, with professional experience at Voya and KPMG. An early AI adopter and lifetime AIAdvantage member.

Portrait of Nick Allen

Nick Allen

Mobile & Backend Engineering

Bio

Leads mobile development for HeyDitto's Kotlin Multiplatform iOS and Android apps and contributes to backend infrastructure. 3+ years in defense contracting, specializing in embedded systems and simulator development.

Portrait of Alan Gothard

Alan Gothard

Head of Marketing & Creative

Bio

Leads brand, content, and marketing. Designs UI and UX for HeyDitto applications. Brings creative and technical experience from Apple alongside audio engineering work and a self-taught jump into web development.

Portrait of Sami Barazanji

Sami Barazanji

Head of Sales & Social Media

Bio

Runs social media strategy end to end and leads sales initiatives. Proven track record in direct-to-consumer online sales. Also the founder of MIBCON.

#The company

Omni Aura LLC builds memory-first AI. We started in 2020 as an electrical engineering senior design project at Auburn University in Auburn, AL, under the name Omni Aura Synths: a hardware synthesizer with an integrated real-time light engine, built by a team of seven students. By the time we formed the LLC in May 2021, we already knew the company would be bigger than synthesizers. Before transformers had a name. When AI meant training character-level RNNs on tweets, we were already sketching multimodal neural networks that could translate between senses.

That curiosity is what led us to HeyDitto, a personal AI that remembers you. It's also why the platform underneath it (AgentFlow, Go libraries, SolidJS devtools) is open source. We fund the work by shipping products on top. It's the Plausible.io model: open source core, paid hosted services, proprietary apps.

Our north star: lower the barriers between people and powerful technology. It's why we moved from hardware to AI. It's why we obsess over developer experience. And it's why we open source the tools we use to build everything else.

#How we got here

Omni Aura didn't start as an AI company. It started as a hardware company that discovered the thing it was building was the layer on top.

"A self-aware instrument (like IMAX experience) that creates smell, sound, light, tactile. OmniSynth experience."

Meeting notes, May 13, 2021

The very first business meeting already sketched something that needed AI to work: an instrument that could react to how you felt and change what you saw, heard, and felt back. We didn't know we were describing agentic AI. We just knew the hardware alone wouldn't get us there.

"GPT-3 assistant 'ditto' becomes the omnisynth assistant. For voice activated tuning synth parameters."

Meeting notes, November 30, 2021

That note predates ChatGPT by a full year. We took the hardware to CES in January 2022 and learned the software and AI framing resonated more than the physical product. We stopped chasing hardware.

Through 2023, Omar built a full smart-home AI stack on a Raspberry Pi, including an Image RAG vision server that went live eight days before OpenAI released GPT-4V. Everything since then is the same 2021 idea catching up to the tools that can run it.

See the full timeline on the home page, or read the long-form version in Before ChatGPT, We Were Already Building Agents.

#Advisers

Alex Patrick

Seed investor and adviser since 2021

Alex invested in Omni Aura during our original period of building music hardware and has continued advising through the transition to AI-native software.