> Markdown version of [/videos/1832-building-and-modernising-apps-with-agentic-ai-julia-kordick?t=1169](https://www.wearedevelopers.com/videos/1832-building-and-modernising-apps-with-agentic-ai-julia-kordick?t=1169). 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).

---

# Building and Modernising Apps with Agentic AI - Julia Kordick

Julia Kordick warns that if agentic AI can replace you, your skills need work. Stop using zero-context prompts. Master spec-driven development to build your developer exoskeleton.

- **Speakers:** [Julia Kordick](https://www.wearedevelopers.com/@julia-kordick)
- **Event:** Coffee With Developers
- **Published:** March 9, 2026
- **Duration:** 38:01
- **URL:** https://www.wearedevelopers.com/videos/1832-building-and-modernising-apps-with-agentic-ai-julia-kordick

## Summary

The shift toward agentic AI is fundamentally changing the software developer's role, transitioning daily workflows from manual coding to managing and directing AI agents. However, many developers struggle with this evolution, relying on weak, zero-context prompts like "fix it" and expecting flawless architectural results. The persistent bottleneck in AI-assisted coding is no longer the generation of raw code, but the lack of mature context engineering and structured technical specification. Organizations without a defined SDLC are adopting AI naively—often granting models unfettered system access—which creates significant security and quality control risks.

To bridge this gap, spec-driven development introduces a structured, iterative loop for managing agentic tools. Using tailored frameworks, developers can quickly generate precise markdown files with uppercase bullet lists—the most effective and explicit format for guiding LLMs. Instead of operating in a reckless "YOLO mode," mature engineering teams integrate these AI-driven specifications directly into their existing quality gates. This managed approach ensures that the human-in-the-loop remains the ultimate validator, preventing AI models from hallucinating unverified logic or drifting from the original project requirements.

Maximizing agentic AI requires maintaining traditional risk-controlling principles, such as breaking large tasks into manageable milestones. Relying on AI to endlessly regenerate a codebase wastes expensive compute tokens and produces massive, unreviewable pull requests that break the proven "divide and conquer" engineering methodology. Instead, forward-thinking teams are building inner-source prompt libraries to standardize success and share optimized context engineering across the organization. Ultimately, AI acts as a developer exoskeleton rather than a wholesale replacement; as Julia Kordick notes, if you are a developer who can be entirely replaced by an agentic AI, "you probably should work on your software engineering skills."

**Keywords:** agentic ai development, spec-driven development, context engineering, developer productivity tools, SDLC guardrails, LLM token optimization, markdown software specifications, inner-source prompt libraries, automated code generation, human-in-the-loop ai, ai pull request management, ai access guardrails, developer workflow modernization, software engineering maturity, ai system prompting

## Chapters

1. **Gathering developer productivity feedback in large organizations** (00:34) — Analyzing customer interactions with enterprise technologies highlights areas for improving software engineering efficiency.
1. **Addressing prompting challenges with spec-driven development projects** (01:10) — Using open tools to simplify agentic coding allows teams to guide code generation with clear specifications.
1. **Transitioning engineering roles toward writing specifications for AI** (02:29) — The evolution of software roles requires developers to quickly learn how to structure work for automation and human-in-the-loop systems.
1. **Setting effective guardrails for enterprise agentic AI adoption** (04:35) — Establishing clear guidelines prevents organizations from blindly trusting AI models with unrestricted system access.
1. **Aligning agentic coding practices with software development lifecycles** (07:18) — Integrating structured processes rather than unmanaged approaches ensures safer and more reliable system code generation.
1. **Navigating friction in spec-driven development and prompt engineering** (10:10) — Balancing dense markdown requirements with concise context helps avoid overwhelming developers and reduces cognitive load during code reviews.
1. **Structuring effective specification formats for AI code agents** (12:41) — Providing short, bulleted lists with explicit uppercase instructions enables autonomous tools to parse technical requirements better.
1. **Iterating on generated code instead of constant regeneration** (14:42) — Revising snippets iteratively saves processing power while preventing the loss of structurally sound early code versions.
1. **Building internal prompt libraries to optimize context engineering** (16:20) — Sharing effective organizational prompts directly reduces token consumption and eliminates generic developer commands.
1. **Maintaining fundamental software engineering skills in an AI era** (19:29) — Relying entirely on generation risks deteriorating an engineer's ability to debug complex integrations or grasp fundamental architectural logic.
1. **Managing oversized pull requests generated by autonomous tools** (23:01) — Breaking large tasks into smaller, manageable milestones prevents unreviewable repo dumps and maintains project stability.
1. **Utilizing open source repositories to test specific agentic workflows** (25:11) — Exploring public specification projects allows teams to test various integrations safely before a full organizational deployment.
1. **Balancing bleeding-edge technology trends with actual customer needs** (27:24) — Recognizing the disparity between social media hype and real-world implementation maturity enables more practical problem-solving for enterprise clients.
1. **Connecting diverse technical communities at developer industry events** (31:01) — Fostering an environment that bridges global vendors with local engineering teams accelerates knowledge sharing and functional networking.
1. **Experimenting with agentic coding as an engineering exoskeleton** (35:58) — Adopting generation tools as supplementary aids rather than full replacements enhances current capabilities through self-driven experimentation.

## Related Moments

- [Shifting developer workloads and realistic AI productivity gains](https://www.wearedevelopers.com/videos/1830-wearedevelopers-live-speculaitions) (from "WeAreDevelopers LIVE - SpeculAItions")
- [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")
- [The limitations of spec-driven development with AI](https://www.wearedevelopers.com/videos/100210-why-optimizing-for-system-comprehension-is-key-to-implementing-ai-for-software-development) (from "Why optimizing for system comprehension is key to implementing AI for software development")
- [Shifting towards specification-driven AI development frameworks](https://www.wearedevelopers.com/videos/100263-the-ai-native-software-team-how-agents-are-rewriting-the-sdlc) (from "The AI-Native Software Team: How Agents Are Rewriting the SDLC")
- [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")
- [Balancing AI tool mandates with developer trust and productivity](https://www.wearedevelopers.com/videos/1365-wearedevelopers-live-the-weekly-developer-show-with-chris-heilmann-and-daniel-cranney) (from " WeAreDevelopers LIVE - the weekly developer show with Chris Heilmann and Daniel Cranney")

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

## 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**
- [Principal Software Engineer, Enterprise AI Platform](https://www.wearedevelopers.com/jobs/ext/1467292-principal-software-engineer-enterprise-ai-platform) at **GitHub**
- [Staff Developer Advocate, GitHub Security Lab](https://www.wearedevelopers.com/jobs/ext/1921051-staff-developer-advocate-github-security-lab) at **GitHub**
- [Principal Product Manager, Agent Platform](https://www.wearedevelopers.com/jobs/ext/277541-principal-product-manager-agent-platform) at **GitHub**
- [AI Full Stack Engineer](https://www.wearedevelopers.com/jobs/ext/1354435-ai-full-stack-engineer) at **Almedia**