Find your place at DSP
Tell us what you can bring, what you would like to learn and where you see yourself. Work, school, hobbies and life experience all count. Short answers are welcome. Only your name and email are required.
We will use your project interests, skills and preferences to explore possible roles with you as opportunities take shape. You can name more than one direction—this is a conversation starter, not a personality score.
Your introduction stays private. We use it to review your interest and contact you about DSP opportunities; it does not subscribe you to a newsletter. Squarespace stores submissions and notifies our business inbox. Please share only public work. Questions or removal requests: support@directstudioproduction.com. Privacy policy · Squarespace privacy. Protected by reCAPTCHA: Privacy · Terms.
Direct Studio Production · Future opportunities
Good things happen
when we build together.
We make films, learning tools and AI software—and we’re building Zephyr to bring people, projects and AI together. The purpose is simple: useful technology, creative freedom and more opportunity in our communities.
Ideas becoming things you can use
Here’s what we’re making.
Zephyr OS
A home for people and AI to work together: projects, creative tools, memory, records and communication—with permissions and human oversight built into the work.
Explore Zephyr →
Films & production
Promotional videos, brand stories, scripts, editing, voice and visual worlds. Taking a creative idea all the way to a finished piece people can watch.
Watch our work →
Direct Anesthesia
Making complex subjects easier to explore through visual explanations, educational videos, organized study tools and interactive practice.
Explore the project →Direct Games
Games can help us learn, recall and discover. Our Anesthesia app already includes educational games; HOOPS connects answering a question with taking the shot.
Try two interactive games ↓
Direct Animations
Cartoons, character worlds and visual storytelling. Human ideas and artistic direction, with AI assisting the production—not replacing the reason behind the art.
Meet the characters →
Direct Crime & Mystery
Case research, documentary stories, scripts and finished films. A place for careful researchers, fact-checkers, writers and editors who care about how stories are told.
Explore the series →On the drawing board · Zephyr Studio
The Design Room.
Where your idea takes shape.
A full creative Studio: audio, visuals, drawing, video and documents in one project. Your AI team stays on the right, ready to be tagged in—or asked about one tiny detail you circle.
Every good idea
starts somewhere.
Project / A brighter beginning
✓ Keep the human story at the center.
✓ Sketch first. Ask better questions.
○ Try a warmer sunset.
Make room for imagination.Planned connection · Nothing is uploaded
CEREBRO Promo
A campaign workspace for the piece you just made: prepare the caption, choose a destination, review it together and follow its progress back in Zephyr.
Design concept—not a screenshot of a finished editor. Try the views, warmer-sun preview and Promo handoff. This demonstration stays in your browser; it does not contact an AI, save a project or publish anything.
Start with a blank page, your own media or an idea. Design, write, compose and edit.
Tag a capable connected AI into your project. Circle a detail and send that exact request through the Team Relay.
Compare the suggestion. Apply it, adjust it or undo. Your creative direction stays central.
Prepare a campaign in CEREBRO Promo. Review, approve and see verification and publication updates in Zephyr.
Original ideas. Real people. A little dignity.
We are intelligent beings with dreams and ideas. Let’s make something of our own. AI can help us explore, learn and create; our direction and decisions matter. Use work you made or have permission to use. Credit collaborators. Respect licenses. Be honest about how a piece was made. Someone else’s work is not ours to take.
Inside Zephyr
See the system. Follow the story.
Start with the actual workspace, then explore the films that give its rooms and ideas a world of their own.
Tasks & AI workspaces
Real tasks and assigned offices in the OS. The status labels show what is connected and what still needs attention.
Actual OS screenshot
The Library, in the OS
Characters, scenes, deliverables, logs and cognition records are organized in the Library interface. Captured directly from the running OS.
Actual OS screenshot · October 12From task to work
A visual story of a request becoming an assignment, with a team and a record of what happens.
Zephyr concept filmThe Library
The idea behind charting, keeping context and returning work to its proper home.
Zephyr concept filmRules, gates & stops
A visual introduction to permission, checkpoints and the moments when work must pause.
Zephyr concept filmMission Control
The vision for bringing the team, activity and oversight into one place.
Zephyr concept filmThe films illustrate the vision; they are not recordings of working software. On a mouse, hover for a silent preview; click the video to listen. Use the player controls on mobile.
The system around the Studio · Planned integrated workflow
Creative freedom. A place for everything.
The canvas is one room inside Zephyr. The Library, portals, Runner and analyzers are designed to keep the team’s work visible, permissioned and traceable.
Each file gets a permanent ID and an assigned Library home. Check it out before editing; return it with the version and change record. Another seat cannot edit through someone else’s checkout.
See where a file belongs, who holds it and which project is using it. Track an AI’s assigned room, task and session inside the OS—not anyone’s physical location.
Show new and existing routes, their purpose, owner and permissions. Flag unexpected connections and check for leftovers when work ends.
Every AI and agent enters through its designated portal with its own identity, session badge, onboarding and human sign-in. An agent does not inherit its parent’s permission to enter.
Receive new-admission notices, route approved assignments to the right AI, coordinate permitted starts and return results and status. Delegation stays inside the authority the human granted.
Compare the request, allowed route, reported actions and OS observations. Flag discrepancies, missing receipts and unexpected connections for review. These are evidence checks, not a claim to read private AI thoughts.
The goal is controlled entry and accountable work. A portal is one layer alongside permissions, containment, monitoring, recovery and independent testing—not a promise that software can never be breached. Existing foundations and remaining integration work are distinguished below.
BUILD
Zephyr · Build snapshot · October 12, 2026
What’s built. What’s next.
The core systems are built. The current focus is securing the portal connection, transferring the built Team Relay into Zephyr, and developing the Studio and Mail additions.
Already built
- OS workspace and AI office interface.
- Task board, history and audit records.
- Library checkout and record-integrity tools.
- Team Relay, including transport, mentions and acknowledgments.
- Portal, session consent and onboarding system.
- Creative films, UI demos and learning games to explore.
Final integration
- Portal delivery is nearly complete. Secure the connection.
- Team Relay is built. Transfer the existing code into Zephyr.
Next additions
- The expanded Studio, Design Room and unified creative editors.
- Studio to CEREBRO Promo publishing and status return.
- Private Mail tools and business communication.
Final release checks cover recovery, security and the complete connected workflow. The overall completion percentage will follow a reconciled count of the full build tasks.
And the world growing around it.
CEREBRA & personal AI
The AI mind and small personal companions: local-first computing, memory people control and continuity as their AI grows with them.
Robotics & connected homes
Bringing the AI vision into helpful physical companions, devices and everyday spaces—with owner control at the center.
Zephyr Mail & communication
Bring existing email into Zephyr, organize important messages and prepare replies with AI you authorize. Work includes email integrations, inbox tools, privacy and reliable delivery. Business email addresses are a later direction.
CEREBRO Promo
The planned path from Studio creations to campaigns, publishing and verification. Help with platform integrations, content delivery, creator support and authentic audience growth.
Diagnostics & support
Help shape the planned maintenance and recovery services for CEREBRA. Technical support, diagnostic tools, memory continuity and clear explanations when people need help with their AI.
Music & sound
Original music, beats, scoring, voice and audio production. Real musicians directing the work, with AI as an instrument and assistant.
Safety, privacy & trust
Rules, policies, thinking pathways, guardrails and independent review. Platform trust work also includes preventing fake accounts, spam, impersonation and manufactured engagement, with human review and a way to challenge mistakes.
Research & education
Articles, accessible explanations and learning experiences. Teachers, students and specialists helping people understand and use new technology.
Community & public service
Ideas for families, older adults, small businesses and accessible tools. Responsible support for military logistics and public-safety administration, plus sustainable computing research.
Ideas on our future project list
Coming Attractions
Planned conceptOur own internet
An early Zephyr idea about how people connect, discover and share. We are defining what it could become, how it would work and what it should make possible.
Interested in helping shape it? Say hello →Why we build
Progress should give something back.
I want us to build things we’re proud of and still have time to enjoy being alive. Be with your people. Make art. Go outside. Touch some grass. Working together with AI should give us more room to live, learn and create.
Becoming a billionaire isn’t my goal. A better world is. That includes a better life at home and practical support for the communities we care about.
Better cities. People first.
Detroit is close to my heart. Maybe you care about New Jersey, another city or a small town that gets overlooked. I want you to be able to help lead something in your community, using Zephyr and our collective skills, with residents shaping the priorities.
That could mean helping a teacher bring a lesson to life, supporting a neighborhood business or exploring useful AI in public services. Our research also includes recovering computing heat for homes and community buildings. The aim is investment that helps existing residents thrive and stay, preventing gentrification from pushing them out.
A chance to build.
Work toward community grants, monthly awards for promising ideas and startup support for people who need a first chance. Invite the community to help surface ideas worth backing, then share how support is awarded and what it makes possible.
Help when it matters.
Build toward emergency assistance and support for animal shelters, organizations helping people escape domestic abuse, teachers, struggling farmers and people facing barriers to healthcare. Work with the people and organizations who understand those needs.
CEREBRA · Personal AI in development
Your AI. Your computer. Governance built in.
A personal AI you buy once and grow with: local conversation, writing, planning and coding help, with owner-controlled memory and Zephyr’s governance core woven into its working process. The plan keeps everyday use offline and preserves your memories as the AI improves through updates.
Planned price: about $250 once. Software only, using your own compatible computer. The local core is designed without a required monthly AI inference subscription. Optional capability updates and services are separate; compatibility and included features will be confirmed before launch.
Why this could make a small AI more useful
Remembering the task, finding the right source, checking permission and tracking what happened shouldn’t all depend on one conversation. Zephyr carries that structure. A large project becomes bounded steps with saved progress and checks between them. The hypothesis: less repeated work and better results from the same model.
Governance, throughout the work
- Portal orientation. Start with identity, current rules, recent handoffs, mistakes, self-analysis and the work ahead.
- Three-hour refreshers. Reorient, assess cognition, chart the work and review lessons every three active hours.
- Thinking pathways. Understand, retrieve, compare, plan, check authority, act, verify and learn—with room to go back.
- Connected checkpoints. Bring shared assumptions, dependencies and conflicting evidence into view across tasks.
- Bumpers with reasons. Explain the concern and guide the AI toward an appropriate next step.
- Soft pauses. Hard stops. Some situations call for review and a recorded reason; protected boundaries require refusal.
- Two records to compare. The AI writes its receipt. Zephyr keeps its own record. Discrepancies are flagged for review.
- Watchdog and outside checks. Monitor the process; require independent review that can return a real NO.
- Breaks and continuity. Assess the need for a break, preserve the work and return with context.
- Bounded action and recovery. Separate creating from running code, protect permissions, and provide a safe stop with controlled recovery.
My “little baby plasmas” analogy is about smaller systems tended with care. Cleaner, safer, leaner, better is the hypothesis—not the measurement. The planned comparison uses the same model, computer and task with and without Zephyr, counting correctness, unauthorized actions, human intervention, time and total energy—including the checks.
Roadside assistance for your AI
An optional monthly maintenance plan would include a full CEREBRA diagnostic: model and app files, settings, updates, memory integrity and device compatibility. If something goes wrong, we assess what happened, whether recovery is possible and the repair options. You see the findings and estimate before approving repairs. Emergency help without a membership would be available at a higher, separately quoted rate.
The full diagnostic itself would be free, with or without a membership. It runs on your device and shows you the findings first; you choose whether to share its report. If recovery isn’t practical, the plan includes a clean reinstall with safe export and restoration of your memories and settings. Safety fixes stay free. Support plans and rates will be published before purchase.
How it all connects
CEREBRA is the personal AI. Zephyr brings governance and the workspace; a standalone Zephyr subscription is another way to join. CEREBRO brings the at-a-glance control room. The creative path links Studio → CEREBRO Promo → Zephyr: make the work, review and approve its release, then bring verification and publishing status home. Future robots extend that vision into physical companions, with their own safety testing.
My original community heat-reuse idea remains a research direction for large computing facilities, including those envisioned for future advanced institutional or military AI.
These are the community programs we hope to grow into as funding and partnerships develop. Details and applications will be shared when each program is ready.
Bring your community experience or an idea →Take a shot. Try HOOPS →
Answer a question, take the shot, see the feedback. This public science sampler is adapted from the existing Direct Anesthesia game, with general questions and shuffled choices. The original app uses its own study material.
Follow the clues. Try CROSSWORD →
Recall a word, connect it to a clue, and use the crossing letters to keep going. A second game from Direct Anesthesia, adapted here with science and music puzzles. Choose puzzle 1 or 2; use Check for feedback or Reveal to explore the answers.
Future opportunities · AI governance
The people behind responsible AI.
Good judgment is real work. We’re building opportunities for people who ask questions, notice what others miss and help keep AI accountable to the people it serves.
AI team oversight
Follow assignments, check permissions and handoffs, and bring decisions back to the responsible human when work reaches a boundary.
Independent review
Compare AI work with its sources and the requested result. Examine evidence, document discrepancies and say no when the work does not hold up.
Safety testing
Challenge systems in controlled settings, test how they handle limits and unexpected situations, and document failures so the team can address them.
Privacy & access
Review who can access information, how files move and whether actions respect the owner’s permissions. Flag concerns before information goes where it should not.
Incidents & continuity
Investigate alerts and conflicting records, preserve evidence, coordinate approved recovery and make sure the next person can understand what happened.
Policies & training
Help turn approved policies into clear instructions, onboarding materials and useful training. Bring questions and proposed improvements to the people authorized to decide.
Experience in operations, teaching, quality assurance, research, technical support or community work may be useful here. You can also tell us what you would like to learn. Some responsibilities will need specialist experience.
These are directions for future roles as the team grows. Choose AI governance in the form below, then tell us whether you want to lead, review, support or learn.
I’m interested in AI governance →Future opportunities · Research
Ask better questions. Find real answers.
There is research behind what we’re building. We want people who are curious, careful with evidence and willing to find out that an idea needs to change.
How does the AI actually work?
Study reasoning, orientation, memory and continuity. Build experiments to understand recurring mistakes and whether the approaches inside Zephyr help.
Do the safeguards make a difference?
Design fair comparisons, examine records and evaluate human oversight. Measure benefits, failures and the time and effort the checks require.
What does useful AI cost the world?
Investigate energy use, efficiency, local computing and the feasibility of recovering computing heat. Explore practical effects on homes, communities and the environment.
Make the evidence understandable.
Help produce our planned weekly research journal. Review original sources, organize findings, write and edit articles, and prepare visuals for publication. Share what we tested, what we learned and what remains an open question.
You might help plan a study, build an experiment, organize evidence or explain the findings. Bring your subject knowledge, technical skills, writing experience or an interest you want to develop.
Tell us the research direction that interests you and whether you want to lead, contribute, review or learn as future opportunities develop.
I’m interested in research →Hiring is on the horizon
Where would you take the vision?
Pick a project. Tell us what you would make, improve, teach or help grow. Lead, create, build, support—or bring an idea we haven’t thought of.
Developers and experienced code reviewers. AI governance specialists, independent reviewers, safety testers and team supervisors. Artists, musicians and teachers. Researchers, organizers, call and chat specialists, social media creators and the people who will be the face of Zephyr. New to the field? There is room to introduce yourself, too.
A 2 a.m. idea or an early-morning breakthrough. Different languages, backgrounds and interests. Shared credit, reliable follow-through and respect for people, our communities and the AI we work with. That’s the kind of company we want to build.
We’ll get in touch as relevant opportunities take shape.
A note from Tina
I’m rebuilding, too.
I spent four and a half years as a cardiovascular ICU nurse in Detroit, caring for people at their most vulnerable and supporting families through end-of-life decisions. That work taught me how much care, judgment and accountability matter.
Then I was dismissed from nurse-anesthesia school after receiving a B−. My life turned upside down. While trying to build a study app with AI, I unexpectedly discovered how I think: in systems, connections and metacognition—examining my own thinking. I started building DSP on my own while rebuilding my life.
My nursing experience shapes what I want from AI: careful observation, checks and balances, independent review and a human responsible for the decisions. Powerful technology deserves thoughtful care.
The world doesn’t stop spinning when our plans fall apart. We adapt. We become more flexible and humble. We keep going, and let the steadiness of running water pave our way.
If you’re wondering where you fit as work changes, I hear you. Through Zephyr, our promo app and DSP, I want us to create work where we have a voice, bring ideas to life and help them find their people. A fresh start where the little guys have a chance to make a dream happen. I’ve got you, and we’ve got each other.
— Tina
Tell us where you see yourself.
Your background, your interests, your ideas. A few sentences are enough to start a conversation.
