Joosep Simm
Developer, Architect & AI Trainer
Telecommunications | Blockchain | AML | Energy | Game Design | Agentic Coding
Six Fridays, senior engineers beside you, your real codebase. You direct the agent, it ships production code.
€2,520/dev · 🇪🇪 Töötukassa covers 80%.
For engineers with 2+ years experience · ~20 per squad
The Shift
Forget the hand-waving about “autonomous systems.” An agent is an LLM that uses tools, runs in a loop until the task is done, and keeps context across iterations.
LLM + Tools + Loop + Context
while(task) { explore → act → verify }Single prompt, single response
prompt → response → doneWhy Claude Code →“After writing my first LLM loop, I couldn't sleep. Once you see it, your perspective changes permanently.”
Kaido Koort
The Bigger Picture
Kaido Koort founded Plepic in 2025 after 23 years in software: 8 teams, 6 products, both agency and client side. He knows what it looks like when efficiency arrives and the teams that made it possible don’t share in it.
Agentic AI changes how code gets written. The next question: why are agencies still billing by the hour? We’re building Scopeful: pricing tied to outcomes, not time.
Meet the Team
Developer, Architect & AI Trainer
Telecommunications | Blockchain | AML | Energy | Game Design | Agentic Coding
Tech Leader & Product Strategist
Led 8 teams | Launched 6 products | Former Head of Dev, Smart-ID | StaffLogic | SurfCast | Lunar Base
Software Engineer & Founder
Founder, Balancing.services | Ex-Sympower | Ex-JRebel | Taught Algorithms at University of Tartu
“The biggest value was a mindset shift: not using AI as a tool, but building architecture around it. An entirely new perspective on applying AI systematically.”
Kristina Krist, Mindworks Industries
“The content was very strong. The biggest value came when I started experimenting on my own project. That’s when real understanding and trust emerged.”
Erich Jagomägis, Bitweb
“It broadens your thinking and shows how to bring AI into your work processes. It gave me a clear direction for where to go next.”
Merilin Sellik, Rethinkers
Common Questions
Agentic coding is a software development approach where an LLM uses tools in a loop to plan, write, test, and refine code autonomously. Unlike single-prompt AI, an agentic system iterates until a goal is achieved, letting developers ship 3–5× faster by delegating repetitive tasks to an AI agent.
ChatGPT responds to a single prompt and stops. An agentic AI runs in a loop: it plans a task, takes actions using tools (like writing files, running tests, or searching code), observes the results, and continues until the job is done. The difference is autonomy over multi-step tasks, not just conversations.
Claude Code is Anthropic’s terminal-based agentic coding tool, released in February 2025. Unlike IDE plugins, it has full system access: read files, run commands, manage git, and integrate with any tool in your pipeline. It’s scriptable end to end, so you extend it with your own commands, skills, and hooks, which is why we built our curriculum around it.
The agentic harness is the structured setup that makes Claude Code work reliably on a real codebase: CLAUDE.md configuration, documentation architecture, brainstorming and scoping skills, the agentic loop, and git worktrees for parallel agent work. The training teaches you to assemble this harness for your specific project.
Senior engineers with 2+ years of professional experience and access to a real codebase. The work agencies, product companies, and AI labs are now hiring for: walking into unfamiliar code, partnering with AI agents, shipping outcomes. Not for learning to code.
The training runs over six Fridays, each a full day (9:00–15:30) via Google Meet. In total the program covers 50 academic hours and is structured so participants can apply what they learn in their own codebases between sessions.
Yes. Participants who complete the program receive a Plepic Certificate of Completion in Agentic Development. The certificate covers the agentic development framework including Claude Code, CLAUDE.md architecture, context management, and the full agentic loop with git worktrees.
Yes. No prior Claude experience is required, only 2+ years of software development experience and access to a real codebase. The first session covers Claude Code setup from scratch, including token economics, commands, and configuration. Teams without prior AI coding tool experience catch up quickly.
We work on your actual codebase, the code you’ll ship Monday. Every session is led by senior engineers with 20+ years each who build production systems with agentic tools. The curriculum was designed in 2025 specifically for Claude Code, not adapted from generic AI content. If you don’t experience at least 3x time savings in typical development tasks after completing the training, we’ll refund your contribution in full. Conditions apply.
€2,520 +VAT per developer. For Estonian employers, Töötukassa reimburses up to 80% of the training invoice including VAT, capped at €2,500 per employee over three years; the employer applies in e-töötukassa before training starts and we prepare the application with them. Plepic koolitus (Plepic OÜ, reg 11275881) is the registered training provider, EHIS majandustegevusteade nr 260952, and the program meets Töötukassa’s minimum of 50 academic hours.
Yes. The training runs online via Google Meet and is open to European teams at the standard price of €2,520 per developer. The Estonian state subsidy is only available for Estonian-registered employers. We run cohorts in English and can accommodate teams across the EU.
A 30-minute call with Kaido. No prep, no slides, no obligation: bring one real task from your backlog and we'll talk through what an agent would do with it.
Book a 30-minute callOpens Google Calendar to pick a time
If you don’t experience at least 3x time savings in typical development tasks after completing the training, we’ll refund your contribution in full. See conditions
Not ready to talk? See the webinar deck →
Curious play. Epic growth.