Research on AI-human collaboration, cooperation structures, economy models, and risk management in the age of terminal AI tools.
Methodology and patterns for effective AI-assisted software development. How to think about, structure, and iterate on code when your pair programmer is an LLM. Reproducibility, prompt engineering, and workflow optimization.
SSH changed the world. With AI, there are many new ways security protocols are evolving. We research how trust, authentication, and encryption adapt when AI agents become network participants.
How AI is transforming classical business models. Token costs, inference economics, build-vs-buy decisions, and the emerging unit economics of AI-native products and services.
Good AI can't stay expensive. It's critically important to think about how to make advanced AI accessible to talented people who want to contribute to the world, regardless of their economic starting point.
From AI-designer to AI-recruiter, from AI-constructor to AI-modelist engineer. Applied research on real-world agent architectures, their limitations, and where they genuinely outperform human workflows.
How is communication between people changing in the AI era? What's shifting in soft skills, leadership, conflict resolution, and collaboration when AI mediates and augments human relationships?
We're looking for people who want to make the world better through technology. Not in a vague, abstract way — but by building real tools that help real people work smarter, learn faster, and collaborate more effectively in the age of AI.
We provide students with three months of hands-on mentorship and access to the most advanced AI tools available today — primarily Claude Code and the broader ecosystem of AI-native development instruments.
Education is completely free. Through our experience, we've found that students who work with us contribute more value to the Institute's research than we invest in their training. A great student is a win for everyone — the student grows, the mentor sharpens their thinking, and the Institute advances its mission. Win-win-win.
Pick a direction that excites you — or propose your own. Here are some examples of things students work on:
No interviews. No resumes. We evaluate what you can do, not what you say you can do. Complete these two tasks to get admitted:
Go to soft-skills.transformerinstitute.org and complete the full assessment:
Submit 3 meaningful pull requests to any of our public repositories. Fix bugs, add features, improve docs, write tests — anything that genuinely helps. We review the quality of your thinking, not the size of the diff.
Repositories accepting contributions:
Send us an email about yourself. No formalities — just be genuine. Tell us:
After we review your email, we'll send you the admission tasks. Complete them at your own pace — there's no deadline.
Apply now ✉ vladimir@transformerinstitute.orgPublished research, tools, and open-source projects from our team.
An open-source diagnostic and training platform for human-centered competencies in an AI-driven workplace. Self-assessment, quiz training, and personalized assignments across 10 skill dimensions — from critical thinking to emotional intelligence.
Research outputs, tools, and experiments across all directions.
| Project | Direction | Description | Status |
|---|---|---|---|
| Soft Skills Check | AI & Human Interaction | Open-source diagnostic platform for 10 core soft skills in the AI workplace | released |
| Claude Code Workflow Patterns | Claude Code Repeat | Documented methodologies for effective AI pair programming — prompt structures, iteration loops, and review flows | active |
| AI Recruiter Agent | AI Agents Usecases | End-to-end recruiting pipeline powered by AI — from screening to interview scheduling | active |
| Gated Knowledge | AI Tools Access | MCP server for AI agents to access auth-gated data sources — Google Drive, BigQuery, Gmail, Slack | active |
| Token Economy Calculator | Unit Economy around AI | Interactive tool for modeling AI product costs — inference pricing, context windows, caching strategies | research |
| AI Agent Security Audit Framework | Secure Protocols Evolution | Methodology for evaluating security risks when AI agents access networks, APIs, and credentials | research |
| Instant Publish | AI Tools Access | Zero-config publishing tool for sharing documents and reports — HTML in, shareable link out | released |
| AI Communication Patterns Study | AI & Human Interaction | Research on how team communication dynamics shift when AI tools mediate collaboration | research |
| Playwright Multi-Session Manager | AI Agents Usecases | Browser automation infrastructure for running parallel AI agent sessions with isolated Chrome profiles | active |
| SSH Trust Models for AI Agents | Secure Protocols Evolution | How authentication and trust chains evolve when non-human actors participate in secure communications | research |