AI Consultant · Educator · Builder
I design and ship agentic AI systems, workflow integrations and full-stack builds, and I teach the teams who will run them.
A career that started in astrophysics and ran through quant finance and design school, with data and code the whole way through. Now I help founders and organizations put AI to real work.
Astrophysics at Columbia (Barnard). Radio astronomy research at NASA Ames, Arecibo and Green Bank.
MA in Statistics from Columbia. Quant and markets roles at Goldman Sachs, Brevan Howard and J.P. Morgan.
MFA in Design & Technology from Parsons. Media-lab work at MIT and The Met. Parsons faculty today.
AI-powered agent pipelines, data archives, internal tools, web and iOS apps.
Client builds, enterprise education and open tools from 2026.
A two-agent AI system: one agent gathers information from sources every day, and a second researches and verifies what it finds. Approved results are published through a review dashboard, so a person makes the final call.
"Demystifying AI" for more than 200 people in Mexico: a bilingual program that gave an entire organization a shared vocabulary for AI, supported by a Spanish-language AI news briefing and reference guides.
Architecture and technology plan for a digital archive of about fifty years of a scholar's work: structured data, a RESTful API and preservation practice. It is designed to be AI-ready, from transcribing audio recordings to searching across the collection, while the archive keeps ownership of its data. The project began in August 2026 and is ongoing.
Custom interactive explainers for corporate teams, used in workshops and client sessions. Examples: the full AI timeline, a benchmark chart, a platform overview across industries, a primer on the "five-layer cake" of the AI economy, or how diffusion models work.
Small-room sessions for decision-makers, built on each team's own problems. One senior leadership team came in with pain points gathered from its staff and left with priorities. In "From chat to built", participants built a working SwiftUI app with Claude Code.
One-on-one work with a single executive. I start by understanding their team, their business and where the friction is, then teach them to implement the workflows and solutions themselves, so the capability stays with them, not with a consultant.
My own framework for launching agent systems quickly. Instead of building each one from scratch, an organization plugs in its sources, rules and publishing destination, and gets a working AI-powered system with deduplication, a human review dashboard and publishing built in.
A curriculum assessment for a graduate design and technology program, asking what students need to learn as AI reshapes the field. The work combined student interviews, rubric-based review and qualitative coding, and ended with recommendations for academic leadership.
I build AI systems end to end: the data pipeline, the model choice, and the interface your team will actually use.
Agentic pipelines, internal tools, dashboards and apps, scoped around a single workflow that should run itself. Every build ships with tests, cost limits and a human review step.
4–8 week buildA clear map of where AI helps your organization, where it doesn't, and what to do first. I cover tool selection, cost comparison, data readiness and a week-one plan.
Short engagementPractical, no-code-required sessions for executives, staff and communities, in English or Spanish, built around the problems your people actually face.
Half-dayFull-daySeriesOne-on-one work with a single executive. I learn the team, the business and the pain points, then teach you to implement the workflows and solutions yourself.
1:1A standing technical partner once systems are live. I monitor, iterate, add capabilities and keep your team current as the models change.
MonthlyI've taught code to designers for more than a decade. Now I teach AI to everyone else.
Parsons School of Design · MFA Design & Technology
I have taught JavaScript, front- and back-end web development, and openFrameworks to BFA and MFA students. Now I teach Data as Material, using data architecture, code and AI to build for the physical and digital world.
Practical AI workshops. Two hours, up to 20 people, no coding required, in person or virtual. Every session includes live demos and hands-on exercises, and people build something of their own before they leave.
Round table taking your team from chatting with AI to personal use cases, built live with your data.
AI session for a senior leadership team, tailored to the specific pain points the organization identified from its own staff.
Large online or in-person workshops on understanding AI in a broad sense — the history, plus case studies for your industry, both positive and negative — helping staff understand where things can go wrong.
Multi-week sessions progressing to fully built and deployed apps, from AI basics to using Claude Code, Terminal, Xcode and other relevant tools for your industry.
Data, privacy and visualization projects from my MFA and media-lab years, incubated at The Met Media Lab, MIT Media Lab, NYC Media Lab and the New Museum's NEW INC.
A browser extension that shows you what your Facebook data says about you, as the machines read it. Developed through NEW INC (2016 cohort) and the NY Media Lab Combine (2016).
Revealing the microbiological world of cities. Honeybees and modified hives collect samples, which are genetically sequenced and turned into 3D web visualizations of each city's microbial neighborhoods. Made with the Playful Systems group, Dr. Chris Mason, Dr. Elizabeth Henaff and others. It grew out of my Parsons thesis on mapping NYC's microbiome.
A web app that uses hip hop to bring younger audiences into The Met's collection. It pairs the Genius (RapGenius) API with The Met's own API across 13 culturally influential rappers.
A design manifesto for the age of AI. It moves from the cosmos down to the everyday, in four parts, and its central question is whether power in technology concentrates or distributes.
The four parts
Regina Flores Mir is an AI consultant and educator whose career spans astrophysics, quantitative finance, and design and technology. She works with founders and corporate teams on data architecture and agentic AI systems, from strategy and workflow automation to full-stack builds. Her work has been incubated at The Met Media Lab, NYC Media Lab and the New Museum's NEW INC, and her MIT Media Lab work went to the 2016 Venice Biennale. She holds degrees from Columbia University and Parsons School of Design, where she teaches in the department of Design and Technology and Data Visualization.