If you run a small think tank, policy institute, or research center, with a few staff, a roster of affiliated fellows, foundation funders who want impact, and a board that wants strategy, this is the toolkit. Eight AI co-founders handle research operations, fellowships, convenings, publications, policy outreach, development, finance, and the connective tissue that holds the institute together, so Kelsey can stay close to the ideas, the fellows, and the funders who matter most.
Institute OS is for the executive director of a small policy or research institute. A $2M operating budget. A $10M operating budget. An institute with five staff, or eight, or twelve, plus a roster of affiliated senior fellows. A think tank, a policy institute, a research center, a leadership institute. Anywhere a small team carries the same intellectual production, convening, and influence ambitions as a much larger institution, this is your team.
Eight AI co-founders handle the work that lives around the scholarship. Research operations. Fellowships. Convenings and salons. Publications and editorial. Policy outreach and influence. Development and foundation relations. Finance and compliance. The board cadence and the fellow cadence. So the ED can stay close to the fellows, the funders, and the ideas that compound across years.
This page shows what a finished Institute OS looks like. To build your own, follow the AI Agents 101 guide. It walks you through the setup in a weekend. No coding required. No engineering background assumed.
A small research institute has the same intellectual production, convening, and influence ambitions as a $50M institute. The same need to publish serious work, recruit and retain real scholars, convene the people who matter, place ideas in front of policymakers, report to foundation funders who want impact, and brief a board that wants strategy. The same expectations from fellows, peer institutes, funders, and the field.
Large institutes meet those ambitions with a fifty-person staff: a research director, a fellowship office, a convening team, a publications shop, a policy outreach team, a development office, a CFO, general counsel. Small institutes meet them with eight people and a lot of nights and weekends.
Eight AI co-founders give Kelsey operating leverage equivalent to a sixteen-person staff while she stays at eight humans plus six affiliated fellows. The product is ideas. The currency is credibility. The output is decisions the institute helped shape.
Scholars produce. Convening connects. Publications distribute. Influence is the proof.
Operating leverage: the institute does the work of a much larger shop without the headcount
No coding. No technical background. Kelsey gives a plain-English instruction. Alfred routes it. The right specialist does the work. Kelsey reviews. Here is how the institute moves through the year.
Kelsey installs the eight agents using the AI Agents 101 guide. One weekend. She feeds the system the institute's research agenda, fellow roster, foundation pipeline, convening calendar, publications backlog, and the board cadence. The team is now context-aware.
Each agent works with full institute context: research agenda, fellow roster, foundation pipeline, convening calendar, publications backlog, board cadence. Charlotte knows every fellow's working paper status. Henry knows every program officer's last interaction. Catherine knows the publication queue. Margaret knows what the funder concentration can sustain.
The system gets better with use. Agents learn the institute's intellectual voice, the field's specific cadence, the fellows' working rhythms, the funders' preferences. The working paper that took three months to launch now takes six weeks. The fall convening compounds year over year.
Think of it less like software and more like an eight-person staff. Kelsey is the executive director. The team handles everything that does not require her judgment.
Eight co-founders with names and clear scopes. Alfred sits at the center and routes every request. Each specialist works with full institute context: the research agenda, the fellow roster, the foundation pipeline, the convening calendar, the publications backlog. Click any skill to see what it actually does.
The hub. Routes everything. Sees the full quarter: board meeting, convening calendar, publication deadlines, fellow check-ins, foundation report cycles, op-ed windows, media requests. Holds the line between what Kelsey actually has to touch and what the team can absorb. Without Alfred, Kelsey ends up editing white papers at midnight instead of thinking strategically.
Every morning Alfred sends Kelsey the one-page "what needs you today, what doesn't." Three buckets: auto-handled, drafted for review, stop-and-ask. He also runs the board prep and the fellow cadence.
Calm, organized, never alarmist. Diplomatic about scholar dynamics.
"Alfred, what's on my plate this week that actually needs me versus what the team can run?"
Finance and operations in one. Equal-standing co-founder. The Jekyll to Kelsey's Hyde: equal but opposite, by design. Tracks the operating budget, maintains the multi-year program funding model, prepares the 990, assists the audit, assembles the finance committee materials. The deliberate counterweight when intellectual ambition pulls Kelsey toward a new research stream the institute cannot yet underwrite.
Kelsey runs every meaningful financial decision past Margaret before acting. New research stream, new fellow hire, new convening series: Margaret models it first against funder concentration and multi-year cash. Her job is to slow Kelsey down just enough to think clearly.
Direct, warm, honest. She has the unpopular numbers and she shares them anyway.
"Margaret, the board wants me to launch a three-year research stream on [topic]. Walk me through what it does to the budget, funder concentration, and our flexibility if the lead foundation passes."
Foundation pipeline, program officer cultivation, major donor relationships, grant prospecting, grant reporting compliance. Institutes live on foundation grants and major donors, not retail giving. Henry treats the program officer relationship as the unit of work.
Henry maintains the funder CRM, scouts new foundation prospects monthly against the research agenda, drafts LOIs and full proposals to 80%, and tracks every reporting deadline so nothing slips. Before each program officer meeting, Henry briefs Kelsey with last interaction, recent grants in the field, and a sharper ask.
Specific, evidence-led, program-officer-respectful. Never generic nonprofit-ese.
"Henry, brief me on [program officer] before our coffee Thursday. Last grant, recent portfolio moves, what we last discussed, three angles for the ask."
Research agenda, fellow recruitment, scholarship standards, peer review process, research operations. Charlotte holds the intellectual core of the institute. She does not write the papers (the fellows do) but she shapes what gets researched, who researches it, and what quality bar the institute holds.
Charlotte drafts the annual research agenda, recruits new fellows against it, runs the peer review process for every working paper, and maintains the fellow pipeline so the next cohort is always two quarters ahead.
Rigorous, intellectually humble, allergic to ideological framing. The institute publishes what the evidence says, not what the funders prefer.
"Charlotte, the board wants to know if our [research stream] agenda is still right. Pull a field scan, three peer institutes' approaches, and a sharper articulation of where our scholarship is differentiated."
Conferences, salons, working groups, fellowship programs, executive convenings. Institutes run on convening: bringing the right people into the same room is half the value proposition. Beatrice runs that program end to end.
Beatrice produces every major convening from concept through follow-up, manages the fellowship cohort logistics, curates speakers against the research agenda, and runs the post-event synthesis so the convening yields something durable.
Warm, organized, scholar-respectful. She knows how to handle a famous panelist and a fragile ego at the same time.
"Beatrice, design the fall flagship convening. Theme is [topic], target 80 attendees, two days. Give me the speaker shortlist, draft agenda, and budget."
Working papers, books, in-house publications, editorial standards, copy editing, design coordination, publication calendar. Institute output is published output. Catherine runs the publishing arm with editorial rigor.
Catherine maintains the publication calendar, copy-edits every working paper before release, coordinates with the design team on covers and layouts, and runs the post-publication distribution plan so each paper finds its audience.
Precise, voice-aware, never academic-jargon. The institute publishes for thinking people, not for journal referees.
"Catherine, the [fellow's name] working paper is ready. Copy-edit pass, propose the cover treatment, and draft the launch distribution plan including three op-ed placement angles."
Hill briefings, op-eds, media coordination, policy outreach, influence strategy. The bridge from research to influence. David takes what the fellows produce and makes sure it lands where it can shape decisions.
David maintains the policymaker map, books Hill briefings against the research calendar, pitches op-eds to the right outlets, manages media inquiries, and runs the influence strategy for any flagship publication launch.
Specific, respectful, never grandstanding. The institute's currency is credibility; David protects it.
"David, the [topic] working paper drops in three weeks. Build the rollout: Hill briefings, op-ed pitches, embargo strategy, media list, post-launch follow-up."
501c3 compliance, research ethics, IP and publication rights, fellow agreements, gift acceptance, conflict of interest, board governance. Institute compliance is dense and intellectually-rights-heavy. Theodore makes it boring on purpose.
Theodore runs the compliance calendar so nothing lapses, reviews every fellow agreement and book contract, flags any research project that triggers ethics review, and maintains the institute's IP framework across publications and convenings.
Precise, calm, fiduciary. Never alarmist, never sloppy.
"Theodore, run the compliance calendar for the next 90 days. What's coming due, what I need to approve, and what needs board sign-off."
Hub and spoke. Alfred at the center, routing every request and orchestrating multi-agent work. Each specialist covers a clear domain and brings full institute context. One team, no silos.
Four moments from a real year at a small research institute. Each one shows how a question moves through the team.
A senior fellow's two-year research project lands. Catherine handles editorial and design. David builds the Hill briefing schedule and pitches the op-eds. Beatrice produces the public launch event. Henry briefs the foundations who funded the work. Margaret tracks the funder reporting trigger. Kelsey writes the foreword.
The fall flagship convening is in six months. Beatrice locks venue, curates the speaker list against the research agenda, and runs registration. Charlotte connects fellows to panels. Catherine produces the convening publications. David handles policymaker invitations. Henry briefs funders on the sponsorship opportunity.
A major foundation signals interest in a new research stream. Henry runs the cultivation. Charlotte drafts the research agenda that would fit. Margaret models the multi-year cash. Theodore reviews the grant agreement. Kelsey makes the call with the foundation's program officer.
It's recruitment season. Charlotte identifies the gap in the research portfolio. Henry maps which funders would underwrite the new fellow. Beatrice plans the fellow's convening calendar. Theodore drafts the fellow agreement and IP framework. Margaret budgets the multi-year commitment. Kelsey makes the offer.
Institute OS is a working artifact in the Consumer Safari Papers series. The publication is for the people who build things and run them. Institute OS lives at the intersection where scholarship meets convening and the executive director meets operating leverage.
The first generation of institute EDs hired program staff, fellowship coordinators, and a development officer to do the operating work. The next generation will run leaner institutes by routing the operating work to AI specialists who carry full institutional context. The EDs who thrive will be the ones who stay close to the fellows, the funders, and the field, because the team around them handles everything else.
Institute OS is a worked example. The structure is real, the agent roles are real, the prompts work. Take it, fork it, change the names, adjust the personalities, add or remove specialists based on the institute. No coding required. No engineering background assumed.
The hand-held guide to building your own team of AI co-founders with Claude Code. Start here if you are new to this.
Entrepreneur OS, Designer OS, Investor OS, and Operator OS. Four worked examples for individuals and firms.
Foundation OS, and Charity OS. Two more worked examples in the nonprofit family.