Software, AI, and physical interaction

Jesse Barrueta

I build software with character. Senior engineer at Apple, with roots in illustration and animation.

Explore my work

Download resume

Engineering meets character

My path into software began with illustration, animation, and Flash. More than 20 years later, I build AI applications and UI systems at Apple, and spend my own time exploring how machines can express attention, hold a conversation, and feel understandable to the people around them.

Southern California, USA

Independent work · 2026

Current experiments

Independent experiments in embodied AI and expressive interaction, exploring how an intelligent machine should behave around people.

Explore 4 experimentsHide experiment details Neural Eyes, persistent AI entities, Jerry, and a tactile music player.

01Research prototype

Neural Eyes

Expression starts with attention.

A learned renderer for independently controlled gaze and expression. I am exploring how a small visual system can communicate character through eye shape, focus, and timing.

Compact PyTorch models and synthetic reference images.
An interactive lab with authored saccades and blinks, recorded behavior traces, and raw model comparisons.
Tools for reviewing illustrated styles and finding visual errors.

The renderer learns the image mapping; the current blink and gaze timing is authored.

Talk about Neural Eyes

02Working service prototype

Enormous Brain entity service

Continuity beyond a conversation.

A reusable service for AI characters whose identity, memories, relationships, and internal state persist independently of a particular model or device. Jerry is one physical embodiment of it.

Persistent entity state and memories, with selective recall using lexical relevance and importance.
Server-side disclosure filtering before information enters model context, with relationship-aware access rules.
Explicit distinctions between verified, probable, and anonymous speaker identity; conversational guesses do not grant access to private context.

Current research directions include semantic retrieval, memory provenance and revision, contradiction handling, and background consolidation.

Talk about the entity service

03Working voice prototype

Jerry

A voice, a character, and a place to live.

A Raspberry Pi talking device connected to the Enormous Brain entity service. It is my testbed for giving a persistent character a voice and a physical presence.

Speech recognition, LLM conversation, and synthesized speech.
Local speaker recognition and push-to-talk interaction.
Experiments with HuskyLens visual sensing and how a character responds to the people around it.

An evolving prototype for embodied AI and human-robot interaction research.

Talk about Jerry

04Hardware exploration

A tactile music player

Give the playhead something you can touch.

A compact, retro-futuristic player concept with physical playback controls and a motorized slider for song position. I am prototyping with Raspberry Pi hardware and exploring the mechanics and feel of a manually movable fader.

Control and mechanism experiments are in progress; the finished player is still a design goal.

Talk about the player

Questions I keep coming back to

What I am exploring

Expressive interaction: How can gaze, timing, sound, and movement make a machine's intent understandable?
AI with continuity: What should a system remember, and how should that context change its next interaction?
Learned interfaces: How far can small learned models take us in generating and responding to visual interfaces?
World models: How can learned models predict the effects of actions and help an embodied system plan what to do next?
Physical controls: Where do tactile feedback and moving parts make an experience more satisfying?

Cortex · Design study

Pick up where you left off.

A concept for carrying project context between a voice device, desktop, and phone, built on the entity service's foundations. I am exploring memory that people can inspect and correct, source records that survive changing models, and explicit permissions for delegated actions.

This is a product and architecture study; the cross-device workflow remains a design goal.

Worked With

Work Experience

Senior Software Engineer

Apple

Full-Time

Sep 2022 - Present

Leading development of AI-enhanced web applications and custom dashboards. Architecting and implementing LLM-powered features for internal tools. Building reusable React component libraries while mentoring engineering teams on AI implementation best practices.

Staff Software Engineer

Carbon Health

Full-Time

Apr 2022 - Sep 2022

Established UI engineering practices and implemented prefix-tree search for patient-facing applications.

Senior Technology Architect & Power Programmer Mentor

Apple, LinkedIn, Infosys

Full-Time

May 2018 - Apr 2022

Architected and delivered four major applications, including the public-facing Sales & Trends platform. Mentored nine engineers while implementing modern tech stacks with GraphQL and Apollo. Led technical interviews for major tech companies and established best practices for multiple teams.

UI Architect

Anthem Blue Cross, UST Global

Full-Time

Jun 2017 - May 2018

Led multiple UI engineering teams while serving as a nationwide advisor on UI architecture. Implemented Test-Driven Development practices and mentored teams on functional programming principles, creating a foundation for scalable applications.

Senior UI Engineer

Beachbody

Full-Time

Sep 2014 - Feb 2017

Created fundamental components for a custom front-end framework and built internal projects using React, Elasticsearch, and MongoDB. Championed Agile processes while working directly with product and marketing executives to deliver high-impact solutions.