Topic
AI & Agents
31 posts on AI & Agents.
Code generation was the easy part
Six engineering leaders on what gets hard after AI writes the code
AI coding agents don't need bigger models. They need better tools.
The same Java migration cost 30K tokens with the right tool and 61M without. Inside the agent tools layer: the LST, local MCP, Prethink, and Trigrep search
Context engineering: why AI coding agents spend most of their tokens reading, not writing
Only 3% of engineers said token cost was their biggest AI coding problem. 42% said agents don't fit their architecture. Why context engineering is the real fix.
Local Moderne MCP: Deterministic Tools for Coding Agents
The Moderne CLI runs a local MCP server that gives coding agents Trigrep search, LST-backed semantic analysis, and deterministic recipe execution, all without leaving the repo
Code quality metrics that AI coding agents can actually use
Cyclomatic complexity, LCOM4, Halstead defect density, test gap risk scores — how PreThink computes and surfaces code quality metrics where agents read them
AI writes code faster than teams can trust it: Quality context is the answer
AI coding agents write code faster than most quality processes were built to handle. The problem isn't the agent — it's that quality intelligence lives on dashboards they never see. This post explains how Moderne Prethink puts code quality metrics directly in the repository, where agents read them automatically before writing a single line.
The first user of your dev tools is no longer a developer
What happens when coding agents use tools designed for humans? Jonathan Schneider on the four tooling gaps costing your agents time and tokens.
The PR is the easy part: What Netflix, Airbnb, and LinkedIn taught me about automated code changes at scale
Engineering leaders from Netflix, Airbnb, and LinkedIn share how they build developer trust in automated code migrations — from shadow validation to PR attribution to migration fatigue.
From grep to Moderne Trigrep: Faster, token-efficient code search for agents and large codebases
Fast, indexed code search for agents and developers. Moderne Trigrep delivers sub-second discovery, lower token usage, and scoped search across large, multi-repo codebases.
Changing the AI context: How Moderne Prethink accelerates coding agents and reduces token use
Introducing Moderne Prethink, a new approach to AI code context that helps coding agents work faster, use fewer tokens, and hallucinate less in large codebases.
Moderne Skills: Agent tools to help AI write and run OpenRewrite recipes
AI-Powered Recipe Authoring with Agent Skills and OpenRewrite | Moderne
AI-driven code maintenance: What works at Meta, Google, and other large enterprises
Why AI succeeds—or fails—at code maintenance in large enterprises. Lessons from Meta, Google, DX, and Moderne on risk, metrics, and scale.
Writing an OpenRewrite recipe with Claude Code: From JBoss to Jetty
See how AI can accelerate writing OpenRewrite recipes for migrations like JBoss to Jetty. Learn how automation improves code modernization at scale.
Implementing large-scale code changes with Moderne
Moderne enables fast, safe, and scalable code remediation across thousands of repos. Learn how teams build trust in automation and modernize with confidence.
Modern, scalable AppSec: Advances in automation and AI security
Learn how to modernize AppSec with automated fixes for legacy code, AI security threats, and compliance needs.
Autonomous testing and the future of developer productivity
Learn how autonomous testing, AI, and intelligent profiling are redefining enterprise codebase management and developer productivity.
Open source auto-refactoring meets AI agent to modernize fintech software at scale
Discover how Moderne and AI agents help fintech teams modernize complex codebases at scale. See why Moderne is presenting with Morgan Stanley at OSFF London 2025.
Tech debt, AI, and the future of software engineering: Lessons from GartnerAPPS 2025
Explore what the 2025 Gartner Application Innovation Summit revealed about modernizing legacy systems, managing tech debt, and navigating the real impact of AI-powered development tools.
Rethinking enterprise code at scale: AI, tech debt, and testability
In this panel recap from Code Remix 2025, learn how engineering leaders are tackling tech debt, boosting testability, and using AI and automation to modernize massive codebases.
Introducing Moderne’s multi-repo AI agent for transforming code at scale
See how Moderne's multi-repo code agent changes the game, enabling developers to understand and evolve entire codebases—not just the repository in front of them.
Overview of OpenRewrite and Moderne
An overview of OpenRewrite and Moderne and how Moderne scales OpenRewrite to enable mass-scale code collaboration and modernization.
What are embeddings, and why are they great for code impact analysis?
Explore how embeddings can be a powerful tool for searching and analyzing large codebases with the help of the Moderne Platform.
AI code search at scale: Finding method invocations with natural language
Discover the power of searching your entire codebase for method invocations using OpenRewrite recipes that leverage a two-phase AI pipeline.
Safe, efficient ways to use AI for mass-scale automated refactoring and analysis
See how Moderne is pioneering the safe and efficient use of AI for automated code refactoring and analysis at mass scale.
Deploying AI LLMs on CPUs: Navigating efficiency and scalability
Learn about our real-world implementation of LLMs on CPUs for mass-scale code refactoring and analysis work within the Moderne Platform.
Announcing AI-assisted refactoring in the Moderne Platform: Where a computer gets AI
Learn how the Moderne Platform has integrated AI LLMs into OpenRewrite recipes to assist refactoring, migration, and impact analysis use cases.
How Moderne engineering built a fast, secure, and cost-effective AI search to assist automated code refactoring
Moderne engineering built an efficient search using AI embeddings. Get the details on the AI LLMs and techniques they employed.
Find personally identifiable information (PII) and secrets exposed by your APIs
Learn how to find personally identifiable information (PII) exposed by APIs and protect sensitive company data.
How to use generative AI for automating code remediation at scale
Learn how to use generative AI for automated code refactoring at scale and why code remediation needs more than an in-line AI assistant.
Moderne achieves SOC 2 Type 2 certification
Moderne is proud to announce that it has achieved the highest level of external audit certification, SOC 2 Type 2 - an extra layer of protection for our users.
What we know so far about generative AI applications to mass refactoring.
Generative AI has enabled new mass refactoring applications, potentially revolutionizing the field. This article provides an overview of the latest research.
No posts in this topic.