> Markdown version of [/videos/1384-compose-the-future-building-agentic-applications-made-simple-with-docker?t=753](https://www.wearedevelopers.com/videos/1384-compose-the-future-building-agentic-applications-made-simple-with-docker?t=753). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Compose the Future: Building Agentic Applications, Made Simple with Docker

Building agentic AI applications is now as simple as running standard containers. See how Docker Compose integrates models as top-level primitives to streamline your multi-LLM architectures.

- **Speakers:** [Jim Clark](https://www.wearedevelopers.com/@jim-clark), [Mark Cavage](https://www.wearedevelopers.com/@mark-cavage), [Tushar Jain](https://www.wearedevelopers.com/@tushar-jain), [Yunong Xiao](https://www.wearedevelopers.com/@yunong-xiao)
- **Event:** World Congress 2025
- **Published:** August 20, 2025
- **Duration:** 50:09
- **URL:** https://www.wearedevelopers.com/videos/1384-compose-the-future-building-agentic-applications-made-simple-with-docker

## Summary

Docker is extending its foundational ethos of making microservices accessible to the emerging landscape of agentic applications, aiming to standardize how developers build, run, and scale multi-LLM architectures. Moving beyond simple task automation, true AI agents require autonomous control loops, secure connections to proprietary data, and goal-oriented programming. To handle this complexity, developers need open, portable toolchains that easily mix cloud-based frontier models with local, open-source small language models (SLMs) to address privacy, latency, and cost constraints.

To bridge the gap between complex AI frameworks and standard developer workflows, models and agents are now integrated directly into Docker Compose as top-level declarative primitives. The new ecosystem allows teams to package models as OCI-compatible artifacts and run them identically to standard containers via the Docker Model Runner. Furthermore, managing the sprawling ecosystem of agent capabilities is simplified by the open-sourced MCP Gateway, which acts as a proxy for the Model Context Protocol. This gateway provides centralized control to scope tool access, filter verbose server logic to optimize agent performance, and inject vital observability into the stack.

The deployment pipeline effectively eliminates infrastructure friction through two major execution environments. For local development, Docker Offload natively integrates remote GPU compute into the CLI, instantly accelerating the inner loop without altering the local developer experience. For production, the exact same `docker-compose.yml` file deploys natively to serverless environments like Google Cloud Run and Azure, abstracting away scalable AI infrastructure. Ultimately, these unified workflows enable developers to focus entirely on context engineering, while leveraging native agentic tools to autonomously audit hardware sizing, orchestrate multi-stage builds, and remediate vulnerabilities.

**Keywords:** docker compose, multi-llm architectures, agentic applications, model context protocol, mcp gateway, docker model runner, docker offload, remote gpu execution, serverless container deployment, google cloud run, hybrid ai infrastructure, context engineering, local slm integration, oci compatible models, open container initiative, autonomous ai control loops

## Chapters

1. **Bridging application containerization and modern AI agent development** (00:05) — The historical shift of containerizing complex environments mirrors the current need to simplify scalable AI agent development.
1. **Defining core characteristics of autonomous AI agent systems** (03:39) — True agents autonomously manage control loops, connect securely to private data, and execute complex goals rather than basic tasks.
1. **Navigating multi-model infrastructures and compound agent swarms** (06:23) — Future agent workflows require mixing foundation models with specialized local variants to balance latency, privacy, and reasoning costs.
1. **Standardizing AI model and tool execution using containers** (08:58) — Packaging models as standard OCI artifacts enables frictionless discovery and execution of open-source resources directly within existing workflows.
1. **Centralizing tool access and observability with MCP Gateway** (12:33) — Utilizing a proxy architecture allows developers to scope server responses, filter tool access, and improve overall agent efficiency.
1. **Declaring full agent stacks structurally with Docker Compose** (14:41) — Treating models and tools as native declarative resources streamlines the composition and instantiation of complex multi-agent architectures.
1. **Architecting cooperating local agents for an ecommerce application** (17:13) — Combining distinct specialized task agents allows developers to parse customer sentiment and process database queries against locally hosted large language models.
1. **Bootstrapping a multi-agent application with generative code assistants** (20:21) — Utilizing inline coding prompt generation securely maps service dependencies and bootstraps execution models for immediate local testing.
1. **Offloading heavy inference workloads automatically to cloud GPUs** (25:33) — Shifting local execution to managed remote GPUs accelerates local development loops without requiring complex network configuration or data migration.
1. **Translating compose stacks natively into serverless cloud environments** (32:45) — Leveraging serverless compatibility ensures identical configuration schemas seamlessly deploy containerized agents as on-demand production services.
1. **Testing high concurrency loads against a serverless agent deployment** (35:05) — Provisioning distributed infrastructure on the fly ensures compound agent chains continuously respond reliably under thousands of concurrent application queries.
1. **Maximizing cloud native capabilities for scaling dynamic AI workloads** (42:22) — Serverless container compute natively supports stateful volumes, auto-scaling instances, and rapid GPU provisioning times for fast application serving.
1. **Automating vulnerability remediation and container builds using AI agents** (45:39) — Deploying dedicated autonomous workspace agents analyzes exposed vulnerabilities and continuously refactors pipeline code to optimize overall image security.

## Related Moments

- [Adopting AI tools for developer container workflows](https://www.wearedevelopers.com/videos/100183-from-build-to-breach-hacking-kubernetes-through-the-supply-chain) (from "From Build to Breach: Hacking Kubernetes Through the Supply Chain")
- [Orchestrating local developer environments with AI tools](https://www.wearedevelopers.com/videos/1392-mcp-mashups-how-ai-agents-are-reviving-the-programmable-web) (from "MCP Mashups: How AI Agents are Reviving the Programmable Web")
- [Integrating generative AI into cloud-native applications](https://www.wearedevelopers.com/videos/950-supercharge-your-cloud-native-applications-with-generative-ai) (from "Supercharge your cloud-native applications with Generative AI")
- [Leveraging agentic capabilities and containerized developer environments](https://www.wearedevelopers.com/videos/1453-10-commandments-for-vibe-coding) (from "10 commandments for vibe coding")
- [Shifting to containerized AI deployment environments](https://www.wearedevelopers.com/videos/1127-using-containers-to-deploy-ai-models-across-our-microscopy-platform) (from "Using Containers to deploy AI Models across our microscopy platform")
- [Practical learnings from deploying containerized AI solutions](https://www.wearedevelopers.com/videos/1127-using-containers-to-deploy-ai-models-across-our-microscopy-platform) (from "Using Containers to deploy AI Models across our microscopy platform")

## Related Articles

- [Building AI Solutions with Rust and Docker](https://www.wearedevelopers.com/magazine/494-building-ai-solutions-with-rust-and-docker)
- [Stephan Gillich - Bringing AI Everywhere](https://www.wearedevelopers.com/magazine/489-stephan-gillich-bringing-ai-everywhere)
- [What is Agentic Programming and Why Should Developers Care?](https://www.wearedevelopers.com/magazine/625-what-is-agentic-programming-and-why-should-developers-care)
- [From Prototype to Production: Build AI Agents with This Free 4-Course Learning Path](https://www.wearedevelopers.com/magazine/655-from-prototype-to-production-build-ai-agents-with-this-free-4-course-learning-path)

## Related Jobs

- [Senior Engineer, Infrastructure Platform](https://www.wearedevelopers.com/jobs/ext/328836-senior-engineer-infrastructure-platform) at **Intercom, Inc.**
- [AI Software Engineer (Germany)](https://www.wearedevelopers.com/jobs/48317-ai-software-engineer-germany) at **Sunhat**
- [Principal Software Engineer, Enterprise AI Platform](https://www.wearedevelopers.com/jobs/ext/1467292-principal-software-engineer-enterprise-ai-platform) at **GitHub**
- [Principal Product Manager, Agent Platform](https://www.wearedevelopers.com/jobs/ext/277541-principal-product-manager-agent-platform) at **GitHub**
- [Staff Software Engineer, Copilot Experiences](https://www.wearedevelopers.com/jobs/ext/164361-staff-software-engineer-copilot-experiences) at **GitHub**
- [Senior AI Agent Software Engineer (Go, Python) (m/f/x)](https://www.wearedevelopers.com/jobs/48277-senior-ai-agent-software-engineer-go-python-m-f-x) at **Dynatrace**