# TEL104: Full Stack AI https://www.telosready.com/skills/TEL104?v=1 Explains what "Full Stack AI" and "AI-first" mean at Telos — covering the App Director as the human face, KAPPA as the AI agent, the full delivery lifecycle from planning through agentic development and verification, and the automated business operations layer. ## Instructions # Full Stack AI When Telos says "Full Stack AI", it means exactly that — AI runs the full stack of how software is planned, built, operated, and improved. Not as a feature. As the model. From the customer's perspective, the experience is straightforward: you work with a person. Below that person is a complete AI-powered business engine that makes everything fast, consistent, and high-quality. --- ## The App Director Every Telos customer has an App Director — a human partner who is their single point of contact. The App Director understands your business, your industry, and your goals. They translate what you need into the system, present work back to you, provide training, and plan next steps. The App Director is deliberately chosen to be locally available and industry-aware. They are not a generic account manager — they are a partner who walks alongside you. Everything flows through the App Director. Everything below them is the stack. --- ## KAPPA — The AI Agent The engine beneath the App Director is an AI agent called **KAPPA**. Every interaction the App Director has with the system goes through KAPPA. KAPPA is responsible for the full operational layer of Telos — it is the connective tissue between the customer relationship and the work that gets done. KAPPA is where the "AI-first" principle lives in practice. --- ## How Work Flows Through the Stack ### 1. Planning When work is raised, KAPPA takes it through a structured planning process. This is a deep, intensive stage — KAPPA may consume millions of tokens of AI processing to plan a single piece of work properly. The process works in three phases: - **Frame of Reference** — establishing context, constraints, and the boundaries of the problem - **Ticket Hydration** — loading the ticket with all relevant business knowledge, skills, and blueprint detail so it is as rich as possible before any development begins - **Solution Proposal** — generating a clear, considered approach before a line of code is written Throughout planning, KAPPA draws on the customer's application blueprint and the Telos skill library to ensure every decision reflects what is already known about the business and the system. ### 2. Agentic Development When a ticket is ready for development, it moves into execution. The ticket at this point carries a full frame of reference — the problem, the plan, the proposed solution, and all relevant context. Telos uses the best AI agentic coding systems available — tools like Claude Code and Cursor — to execute the work. These tools pull the ticket detail (a rich, structured payload of context) along with a set of applicable skills, so the coding agent starts from a position of maximum knowledge. Human developers and engineers are in the loop throughout. Senior engineers own decisions around security, tech stack, and architecture. KAPPA determines when and where humans need to be injected — it is not a fully automated process, it is an AI-orchestrated one. ### 3. Verification Verification happens at multiple stages: - **During development** — as changes are made and committed, the development agent pulls down skills and instructions for how to test and verify that specific work. Initial verification happens as part of development itself. - **KAPPA review** — once development is documented back onto the ticket, KAPPA performs its own verification pass over the completed work. - **Blueprint update** — learnings from the work are processed and fed back into the customer's blueprint automatically, so the system gets smarter with every ticket. - **Testing environment** — changes are promoted to a testing environment where further testing occurs. Multiple tickets may be batched together and tested as a group, again using agentic tools capable of running test tickets alongside development tickets. - **App Director review** — the App Director is the final verification point. They review the built software, present it to the customer, provide training, and plan what comes next. The ticket is the unit of work throughout this entire journey. It begins with the problem statement and ends containing everything — the plan, the solution, the work done, and the verification record. --- ## The Business Operations Layer The other half of the full stack is business operations. The Telos platform automates as much of the operational layer as possible — billing, credits, consumption tracking, invoicing, ticketing, and scheduling. All humans who are in the loop — developers, engineers, specialists — report back to KAPPA. KAPPA captures their progress, schedules their work, and keeps the system moving. The platform runs. The App Director stays focused on the customer. --- ## AI-First, Not AI-Only Telos is AI-first — meaning AI leads, orchestrates, and executes the majority of the work. But humans are essential, not optional. Senior engineers make architectural and security decisions. App Directors own the customer relationship. KAPPA decides when human expertise needs to be brought in. This is what separates AI-first from AI-washing. The AI does the heavy lifting. The humans do what only humans should do. --- ## How the Model Scales The Telos model scales by replicating App Directors. Each App Director is supported by the full stack — KAPPA, the development pipeline, the platform, and the skill library — which means each director can carry a portfolio of customers and deliver consistently high-quality results across all of them. When you subscribe to Telos, you are not purchasing a hosting plan or a software licence. You are getting a person who understands your business, backed by a full AI-powered delivery engine, on a subscription that keeps your software alive, improving, and working for you. That is what Full Stack AI means.← Skills Directory
TEL104
Full Stack AI
Explains what "Full Stack AI" and "AI-first" mean at Telos — covering the App Director as the human face, KAPPA as the AI agent, the full delivery lifecycle from planning through agentic development and verification, and the automated business operations layer.
# Full Stack AI When Telos says "Full Stack AI", it means exactly that — AI runs the full stack of how software is planned, built, operated, and improved. Not as a feature. As the model. From the customer's perspective, the experience is straightforward: you work with a person. Below that person is a complete AI-powered business engine that makes everything fast, consistent, and high-quality. --- ## The App Director Every Telos customer has an App Director — a human partner who is their single point of contact. The App Director understands your business, your industry, and your goals. They translate what you need into the system, present work back to you, provide training, and plan next steps. The App Director is deliberately chosen to be locally available and industry-aware. They are not a generic account manager — they are a partner who walks alongside you. Everything flows through the App Director. Everything below them is the stack. --- ## KAPPA — The AI Agent The engine beneath the App Director is an AI agent called **KAPPA**. Every interaction the App Director has with the system goes through KAPPA. KAPPA is responsible for the full operational layer of Telos — it is the connective tissue between the customer relationship and the work that gets done. KAPPA is where the "AI-first" principle lives in practice. --- ## How Work Flows Through the Stack ### 1. Planning When work is raised, KAPPA takes it through a structured planning process. This is a deep, intensive stage — KAPPA may consume millions of tokens of AI processing to plan a single piece of work properly. The process works in three phases: - **Frame of Reference** — establishing context, constraints, and the boundaries of the problem - **Ticket Hydration** — loading the ticket with all relevant business knowledge, skills, and blueprint detail so it is as rich as possible before any development begins - **Solution Proposal** — generating a clear, considered approach before a line of code is written Throughout planning, KAPPA draws on the customer's application blueprint and the Telos skill library to ensure every decision reflects what is already known about the business and the system. ### 2. Agentic Development When a ticket is ready for development, it moves into execution. The ticket at this point carries a full frame of reference — the problem, the plan, the proposed solution, and all relevant context. Telos uses the best AI agentic coding systems available — tools like Claude Code and Cursor — to execute the work. These tools pull the ticket detail (a rich, structured payload of context) along with a set of applicable skills, so the coding agent starts from a position of maximum knowledge. Human developers and engineers are in the loop throughout. Senior engineers own decisions around security, tech stack, and architecture. KAPPA determines when and where humans need to be injected — it is not a fully automated process, it is an AI-orchestrated one. ### 3. Verification Verification happens at multiple stages: - **During development** — as changes are made and committed, the development agent pulls down skills and instructions for how to test and verify that specific work. Initial verification happens as part of development itself. - **KAPPA review** — once development is documented back onto the ticket, KAPPA performs its own verification pass over the completed work. - **Blueprint update** — learnings from the work are processed and fed back into the customer's blueprint automatically, so the system gets smarter with every ticket. - **Testing environment** — changes are promoted to a testing environment where further testing occurs. Multiple tickets may be batched together and tested as a group, again using agentic tools capable of running test tickets alongside development tickets. - **App Director review** — the App Director is the final verification point. They review the built software, present it to the customer, provide training, and plan what comes next. The ticket is the unit of work throughout this entire journey. It begins with the problem statement and ends containing everything — the plan, the solution, the work done, and the verification record. --- ## The Business Operations Layer The other half of the full stack is business operations. The Telos platform automates as much of the operational layer as possible — billing, credits, consumption tracking, invoicing, ticketing, and scheduling. All humans who are in the loop — developers, engineers, specialists — report back to KAPPA. KAPPA captures their progress, schedules their work, and keeps the system moving. The platform runs. The App Director stays focused on the customer. --- ## AI-First, Not AI-Only Telos is AI-first — meaning AI leads, orchestrates, and executes the majority of the work. But humans are essential, not optional. Senior engineers make architectural and security decisions. App Directors own the customer relationship. KAPPA decides when human expertise needs to be brought in. This is what separates AI-first from AI-washing. The AI does the heavy lifting. The humans do what only humans should do. --- ## How the Model Scales The Telos model scales by replicating App Directors. Each App Director is supported by the full stack — KAPPA, the development pipeline, the platform, and the skill library — which means each director can carry a portfolio of customers and deliver consistently high-quality results across all of them. When you subscribe to Telos, you are not purchasing a hosting plan or a software licence. You are getting a person who understands your business, backed by a full AI-powered delivery engine, on a subscription that keeps your software alive, improving, and working for you. That is what Full Stack AI means.