> Markdown version of [/videos/100624-agentic-software-engineering](https://www.wearedevelopers.com/videos/100624-agentic-software-engineering). 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).

---

# Agentic Software Engineering

Dr Ralf S. Engelschall argues that enterprise AI requires strict engineering constraints to survive. Discover how his open-source toolkit uses LLMs to proactively challenge and refine software architecture.

- **Speakers:** Dr. Ralf S. Engelschall, [Christian Heilmann](https://www.wearedevelopers.com/@chris-heilmann), [Daniel Cranney](https://www.wearedevelopers.com/@daniel-cranney)
- **Event:** Coffee With Developers
- **Published:** October 23, 2026
- **Duration:** 45:49
- **URL:** https://www.wearedevelopers.com/videos/100624-agentic-software-engineering
- **Thumbnail URL:** https://customer-goifxnzhrq3d0d54.cloudflarestream.com/df2861f73bd62dfbe76998ddc3b2a5a3/thumbnails/thumbnail.jpg?time=697200ms

## Summary

**Enterprise AI Adoption:** The rapid shift toward agentic AI coding presents a deep conflict for enterprise software teams: balancing the speed of AI-generated vibe coding with the rigorous compliance, traceability, and risk mitigation required in regulated markets. While raw manual coding is declining, traditional software engineering principles remain critical for systems demanding decades-long lifespans. Furthermore, European enterprises face an immediate data sovereignty challenge, as maintaining a competitive edge currently mandates relying on frontier US models like Anthropic's Claude and OpenAI. **Agentic Software Engineering:** To bridge this gap, the open-source toolkit Agentic Software Engineering (ASE) fuses classical engineering structures with modern AI capabilities. Acting as an intelligent harness for platforms like Claude Code or GitHub Copilot, ASE introduces formal requirements tracking, markdown-based schema linking, and task-based development pipelines directly into the developer workflow. Instead of merely prompting for arbitrary code generation, developers engage in brief engineering, heavily constraining and guiding the AI. **Methodology and Application:** A standout workflow within this framework is grilling, where the LLM proactively interviews the developer to identify underspecified architecture or domain logic before writing a single line of code, drastically reducing post-generation adjustments. The toolkit also features an Advocatus Diaboli skill to logically challenge architectural decisions, demonstrating how AI can evaluate system design beyond mere code synthesis. By integrating these rigorous, Unix-friendly engineering practices with frontier LLMs, organizations can safely migrate legacy systems like COBOL while utilizing modern architectures like TypeScript, empowering teams to leverage AI without sacrificing enterprise-grade reliability.

**Keywords:** agentic software engineering, AI code generation compliance, frontier LLM adoption, vibe coding limitations, enterprise risk mitigation, data sovereignty challenges, brief engineering workflows, legacy code migration, COBOL transformation, software requirements engineering, AI architectural validation, github copilot integration, claude code workflows, typescript development, open source AI toolkits

## Chapters

1. **Introduction to MSG Group and enterprise software** (00:01) — MSG Group focuses on building large-scale business information systems and navigating industry changes.
1. **Navigating AI transformations in enterprise software** (01:16) — Integrating AI features into legacy enterprise products and developer workflows presents unique challenges for large organizations.
1. **Data sovereignty and the reliance on frontier models** (03:55) — European companies are currently forced to use US-based frontier AI models despite growing data sovereignty concerns.
1. **Token optimization and the reality of open weights** (09:22) — The transition from token maximization to local vector caching highlights the distinction between open weights and true open source.
1. **Balancing agentic AI code generation with software engineering** (12:03) — Distinguishing between throwaway AI-generated convenience apps and compliant, long-lasting enterprise software engineering is crucial.
1. **The convenience and vendor lock-in of AI coding platforms** (16:37) — Modern AI platforms repeat historical patterns of low-code tools by offering rapid development at the cost of vendor lock-in.
1. **Introducing agentic software engineering for structured workflows** (22:52) — Merging traditional software engineering processes like requirements tracing with agentic AI coding tools creates a robust workflow.
1. **Exploring the flexible agentic software engineering toolkit** (27:25) — The ASE toolkit provides modular features like Kanban boards and devil's advocate evaluations for software architects.
1. **Improving AI code generation through relentless project grilling** (31:41) — Interrogating and clarifying requirements before implementation minimizes post-generation adjustments in AI coding workflows.
1. **Web development using AI brief engineering techniques** (36:35) — Combining AI constraints and human briefing generates structured code for web frameworks like Astro and Tailwind.
1. **Open source contributions and real-world toolkit usage** (39:37) — The ASE open source project offers practical applications for internal enterprise teams and external contributors.
1. **Legacy code migration and programming language preferences** (42:43) — The focus on COBOL for banking systems highlights the ongoing need for legacy migration alongside modern TypeScript adoption.

## Related Moments

- [Introduction to agentic AI engineering and code scanning](https://www.wearedevelopers.com/videos/100448-don-t-kill-my-vibes-simple-steps-to-stay-secure-when-vibe-coding) (from "Don’t kill my Vibes - Simple Steps to Stay Secure when Vibe Coding")
  - **Thumbnail URL:** https://customer-goifxnzhrq3d0d54.cloudflarestream.com/24488ef1e000cfbf47226082fd7c1ab5/thumbnails/thumbnail.jpg?time=97581ms
- [Shifting developer workloads and realistic AI productivity gains](https://www.wearedevelopers.com/videos/1830-wearedevelopers-live-speculaitions) (from "WeAreDevelopers LIVE - SpeculAItions")
  - **Thumbnail URL:** https://customer-goifxnzhrq3d0d54.cloudflarestream.com/2cdb4226b123a771ffa9261c5605d057/thumbnails/thumbnail.jpg?time=202197ms
- [Rise of AI tools and agentic engineering speed](https://www.wearedevelopers.com/videos/100581-valven-intelligence-for-ai-driven-software-lifecycle) (from "Valven: Intelligence for AI-Driven Software Lifecycle")
  - **Thumbnail URL:** https://customer-goifxnzhrq3d0d54.cloudflarestream.com/c8b119816f799411de68355185dec428/thumbnails/thumbnail.jpg?time=82722ms
- [The evolution of AI programming and agentic workflows](https://www.wearedevelopers.com/videos/100032-under-the-hood-of-building-on-lovable) (from "Under the Hood of Building on Lovable")
  - **Thumbnail URL:** https://customer-goifxnzhrq3d0d54.cloudflarestream.com/938ad01fcf12aa00515a43075be1a319/thumbnails/thumbnail.jpg?time=66750ms
- [Managing organizational change for AI developer productivity tools](https://www.wearedevelopers.com/videos/1457-exploring-llms-across-clouds) (from "Exploring LLMs across clouds")
  - **Thumbnail URL:** https://customer-goifxnzhrq3d0d54.cloudflarestream.com/de9dd96eaedcb46a59491acc99bf7155/thumbnails/thumbnail.jpg?time=30000ms
- [Motivations for adopting AI to enhance developer productivity](https://www.wearedevelopers.com/videos/1266-navigating-the-ai-revolution-in-software-development) (from "Navigating the AI Revolution in Software Development")
  - **Thumbnail URL:** https://customer-goifxnzhrq3d0d54.cloudflarestream.com/0426340daa034ef3c93e81f4dce2b3df/thumbnails/thumbnail.jpg?time=318015ms

## Related Articles

- [What is Agentic Programming and Why Should Developers Care?](https://www.wearedevelopers.com/magazine/625-what-is-agentic-programming-and-why-should-developers-care)
- [Transforming Software Development: The Role of AI and Developer Tools](https://www.wearedevelopers.com/magazine/527-transforming-software-development-the-role-of-ai-and-developer-tools)
- [MLOps And AI Driven Development](https://www.wearedevelopers.com/magazine/82-mlops-and-ai-driven-development)
- [Exploring AI: Opportunities and Risks for Developers](https://www.wearedevelopers.com/magazine/522-exploring-ai-opportunities-and-risks-for-developers)

## Related Jobs

- [AI Software Engineer (Germany)](https://www.wearedevelopers.com/jobs/48317-ai-software-engineer-germany) at **Sunhat**
- [Head of Agentic AI / Lead AI Engineer](https://www.wearedevelopers.com/jobs/48488-head-of-agentic-ai-lead-ai-engineer) at **1st solution consulting gmbh**
- [Senior AI/ML Engineer](https://www.wearedevelopers.com/jobs/48352-senior-ai-ml-engineer) at **PagerDuty**
- [Staff Software Engineer, Agentic Platform](https://www.wearedevelopers.com/jobs/48464-staff-software-engineer-agentic-platform) at **Docker, Inc.**
- [Senior Agentic AI Engineer (m/f/d)](https://www.wearedevelopers.com/jobs/48489-senior-agentic-ai-engineer-m-f-d) at **1st solution consulting gmbh**
- [Senior AI Platform Engineer](https://www.wearedevelopers.com/jobs/ext/3709265-senior-ai-platform-engineer) at **ARM**