Developer Productivity Using AI Tools and Services - Ryan J Salva

The latest DORA report reveals a surprising trend. AI adoption correlates with a 7.2% drop in delivery stability.

#1about 4 minutes

The diverse ways AI assists developers today

AI tools boost productivity through predictive text, conversational chat for code understanding, "vibe coding" for rapid prototyping, and automation agents for complex tasks.

#2about 5 minutes

The risk of prioritizing speed over code quality

The DORA report shows AI adoption correlates with a regression in delivery stability, highlighting the need to shift focus from writing more code to writing better code.

#3about 3 minutes

Using AI to improve code quality and reduce tech debt

AI can improve code quality by generating documentation that reflects the current state of a codebase and can help manage technical debt by automating library and SDK migrations.

#4about 5 minutes

Keeping AI models current with the latest code changes

Retrieval-Augmented Generation (RAG) and techniques like Search RAG help AI models provide up-to-date code suggestions by referencing the latest documentation and code samples.

#5about 6 minutes

Creating team-specific coding standards with AI

AI can learn a team's unique coding patterns and architectural preferences by analyzing code review feedback, enabling it to generate code that adheres to internal standards.

#6about 5 minutes

Ensuring transparency in AI-generated code suggestions

Current AI tools provide transparency through citations and footnotes, but the future will involve developers asking models to explain their reasoning rather than reviewing every line of code.

#7about 5 minutes

How AI is changing the role of the software developer

Developers are shifting from writing code to writing detailed natural language requirements, elevating their role to be more architectural and democratizing development for other team members.

#8about 6 minutes

Using AI as a learning tool for developers

Rather than replacing junior developers, AI assistants can act as a coach or sounding board, helping developers learn new concepts and complete tasks more independently.

#9about 5 minutes

Modernizing legacy codebases like COBOL with AI

AI tools can accelerate the modernization of legacy systems by assisting with code translation, re-architecture, and dependency analysis for complex migration projects.

#10about 6 minutes

Mitigating the security risks of AI-generated code

Security risks like malicious packages and data leaks can be managed by using static analysis tools and running AI assistants within sandboxed, cloud-hosted development environments.

#11about 7 minutes

Why new AI-native IDEs are emerging

The rise of new IDEs is driven by the need for different extensibility points and user experiences tailored to prompt-driven, agentic workflows that traditional editors don't support.

#12about 1 minute

Why operations is the next frontier for AI

The next major opportunity for AI in developer tooling is in operations (Ops), an untapped area that includes log analysis, incident response, and infrastructure design.