> Markdown version of [/videos/100332-software-that-fixes-itself?t=731](https://www.wearedevelopers.com/videos/100332-software-that-fixes-itself?t=731). 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).

---

# Software That Fixes Itself

Milin Desai and Rodrigue Schäfer warn that AI-assisted coding produces triple the bugs. Are you prepared? See how self-healing software automates incident resolution to prevent severe developer burnout.

- **Speakers:** [Milin Desai](https://www.wearedevelopers.com/@milin-desai), [Rodrigue Schäfer](https://www.wearedevelopers.com/@rodrigue-schafer)
- **Event:** World Congress 2026 Europe
- **Published:** July 10, 2026
- **Duration:** 28:38
- **URL:** https://www.wearedevelopers.com/videos/100332-software-that-fixes-itself

## Summary

Exponential growth in AI-assisted code generation threatens to overwhelm downstream operations. As Sentry CEO Milin Desai and Delivery Hero VP Rodrigue Schäfer discuss, shipping three times as much code logically results in a proportional increase in bugs. To prevent developer burnout and handle this increased volume, technical teams must fundamentally automate bug resolution, shifting from purely human-centric incident response to what the industry is calling self-healing software.

Delivery Hero tackled this pre-production bottleneck by building HeroGen, an autonomous agentic delivery system handling tasks from technical debt removal to security remediation. Rather than allowing agents to loosely pull data via MCP servers, HeroGen utilizes a deterministic harness that directly pushes precise CI signals and strict state transitions to the agent, minimizing token waste and hallucination. A standout innovation in their workflow is a council of agents for code review. Because AI models inherently rate their own outputs too favorably, HeroGen employs adversarial models to review generated code, helping it achieve an impressive 85% merge rate across roughly 200 daily pull requests.

As pre-production coding accelerates, the delivery bottleneck inevitably shifts to testing, deployment, and incident response. Schäfer outlines a tiered roadmap for self-healing in production, starting with AI-driven root cause analysis and progressing to automated runbooks for known errors, culminating in autonomous execution of low-risk mitigations with strict human authorization for state-changing actions. Ultimately, AI functions as a new, higher-level abstraction layer—much like the shift from punch cards to high-level programming. This evolution transforms engineers from strict implementers into system supervisors, increasing the absolute value of deep architectural and systems knowledge.

**Keywords:** self-healing software, autonomous agentic delivery, AI incident response, deterministic agent architecture, automated root cause analysis, adversarial code review, automated PR generation, AI code generation bottlenecks, enterprise software production, security vulnerability remediation, on-call engineer burnout, low-risk system mitigation, LLM council of agents, human-in-the-loop authorization, AI software delivery lifecycle, MCP server alternatives

## Chapters

1. **Introduction to software reliability at Delivery Hero scale** (00:44) — An overview of Sentry's mission and the immense operational scale of the Delivery Hero platform.
1. **Managing the operational impact of AI-generated code volume** (02:48) — How the rapid increase in code generation creates a parallel rise in production application incidents.
1. **Automating routine pre-production tasks with the HeroGen agent** (04:57) — Delivery Hero's autonomous agent handles routine implementation tickets to free up developers for architecture work.
1. **Designing deterministic agentic systems for controlled pull requests** (09:48) — Supplying explicitly controlled context to agents prevents token waste and guides accurate code fixes.
1. **Using a council of adversarial agents for code review** (12:11) — Deploying multiple models with distinct perspectives overcomes the analytical blind spots of self-evaluating AI.
1. **Shifting software delivery bottlenecks to operations and incident response** (14:00) — Accelerating upstream coding shifts system pressure toward downstream deployment verification and stressful on-call environments.
1. **The step-by-step roadmap for self-healing software in production** (17:09) — Safely scaling production automation requires progressing from basic root cause analysis to low-risk autonomous mitigation.
1. **Defining success for fully agentic software delivery lifecycles** (22:15) — A mature self-healing lifecycle involves scalable root cause identification alongside automated canary deployment rollbacks.
1. **How natural language interfaces redefine the software engineering role** (25:33) — Raising the computer science abstraction layer to human language empowers engineers to build at greater scale.

## Related Moments

- [Navigating developer bottlenecks and human accountability](https://www.wearedevelopers.com/videos/100265-fireside-chat-in-conversation-with-werner-vogels-cto-of-amazon-com) (from "Fireside Chat - In conversation with Werner Vogels, CTO of Amazon.com")
- [Shifting developer workloads and realistic AI productivity gains](https://www.wearedevelopers.com/videos/1830-wearedevelopers-live-speculaitions) (from "WeAreDevelopers LIVE - SpeculAItions")
- [The future of artificial intelligence in platform engineering operations](https://www.wearedevelopers.com/videos/1946-from-pipelines-to-platforms-how-devops-automation-becomes-a-force-multiplier-at-scale) (from "From Pipelines to Platforms: How DevOps Automation Becomes a Force Multiplier at Scale")
- [Balancing developer autonomy with the adoption of coding agents](https://www.wearedevelopers.com/videos/100198-the-last-mile-of-ai-from-prototype-to-production) (from "The Last Mile of AI: From Prototype to Production")
- [Impact of AI on the software development lifecycle](https://www.wearedevelopers.com/videos/100086-unlocking-the-ai-black-box-building-trust-in-the-era-of-agentic-production) (from "Unlocking the AI Black Box: Building Trust in the Era of Agentic Production")
- [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")

## 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)
- [What is Software Engineering in the Age of AI?](https://www.wearedevelopers.com/magazine/640-what-is-software-engineering-in-the-age-of-ai)
- [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)
- [Exploring AI: Opportunities and Risks for Developers](https://www.wearedevelopers.com/magazine/522-exploring-ai-opportunities-and-risks-for-developers)

## Related Jobs

- [Staff Software Engineer, Copilot Experiences](https://www.wearedevelopers.com/jobs/ext/164361-staff-software-engineer-copilot-experiences) at **GitHub**
- [AI Software Engineer (Germany)](https://www.wearedevelopers.com/jobs/48317-ai-software-engineer-germany) at **Sunhat**
- [Tribe Lead - ( Software) Engineering Centre of Excllence](https://www.wearedevelopers.com/jobs/ext/1475530-tribe-lead-software-engineering-centre-of-excllence) at **SD Worx**
- [Principal Product Manager, Agent Platform](https://www.wearedevelopers.com/jobs/ext/277541-principal-product-manager-agent-platform) at **GitHub**
- [Senior Software Engineer, Enterprise Products](https://www.wearedevelopers.com/jobs/ext/1841248-senior-software-engineer-enterprise-products) at **GitHub**
- [Principal Software Engineer, Enterprise AI Platform](https://www.wearedevelopers.com/jobs/ext/1467292-principal-software-engineer-enterprise-ai-platform) at **GitHub**