AI Code Suggestions in Snowsight are now GA. Under the hood, our Snowflake AI Research team moved SQL autocomplete from a 30B-A3B MoE to a 4B dense model specialized for Snowflake SQL. • Acceptance: 17.80% → 26.35% • Latency: Down 71% The full research:

AI can write more code than any team can review by hand, and a pull request can look fine while hiding a security issue or missing a requirement. In our new short course, AI Code Review, built in collaboration with @QodoAI and taught by @nnennahacks, you'll learn the practices that make AI code review effective: review before you open a pull request, give the reviewer full context about your codebase, and triage findings by risk. Then you'll build your own review agent, from a context engine that finds the right code to a team of specialized reviewers. Enroll for free:

Hot take: AI code reviews are becoming another form of technical debt. We're generating more code, asking AI to review more code, and then asking another AI to fix what the first AI found. At what point do we stop creating problems just to automate solving them?

Internet Computer’s AI Code Generator is Unreal According to the founder of ICP (@Dfinity), the AI coding assistant @Caffeineai has generated 15 billion lines of code. A large portion of that was generated for Motoko, the native language of the Internet Computer’s autonomous agents. With this spike in development, “Open SaaS” and “NextGen AIware” will be rolled out faster than ever, positioning $ICP as a key execution layer for agentic services. Motoko’s development enables developers to create autonomous applications running on the blockchain, free from any cloud dependency.

Before an AI agent can actually get work done, it needs a whole new stack: an identity, and the ability to search, to read documents, to review code, and to pay. Tomorrow in San Francisco, @GoKiteAI shares the stage at Agent Experience Demo Night, hosted by Sapient, with a lineup of teams building the rest of that stack: ▷ @auth0: the identity and login layer, acquired by Okta ▷ @Docusign (Nasdaq: DOCU): the global standard for e-signature and agreements ▷ @Box (NYSE: BOX): enterprise content and documents in the cloud ▷ @nebiusai (Nasdaq: NBIS): AI cloud and GPU infrastructure for training and inference ▷ @coderabbitai: AI code review on every pull request ▷ @tavilyai: the search API built for AI agents Our Co-Founder & CTO @scottshics will demo Kite's piece: a verifiable identity for every agent, and the ability to spend within limits you set. 📅 July 7th (Tue), 5:00-9:00 PM PDT 📍 Okta, 100 1st St, San Francisco Register (host approval required): See you there. 🪁

Burning tokens on AI code review? 🔥 This tool turns your codebase into a structural graph so the AI reads only relevant files — cutting tokens by a median of 82x. Title: tirth8205/code-review-graph URL: 🔥 Overview A local-first code intelligence tool that persists a structural map (graph) of your codebase via Tree-sitter parsing. It lets AI assistants review by reading only contextually relevant files instead of the whole repository. ❓ Challenges Solved AI code review tools waste tokens by re-reading large parts of the codebase on every review. ・In large monorepos, context bloats and both cost and latency worsen ・Analyzing change impact required scanning the entire project 💡 Methodology & How It Works A three-stage pipeline. ・Parsing: Tree-sitter builds ASTs, extracting functions, classes, imports, and call relationships ・Graph storage: nodes and edges persist in SQLite (no external database) ・Analysis: on changes, blast-radius analysis traces affected callers, dependents, and tests, returning minimal context It supports many languages, incremental updates under 2 seconds, MCP integration (30 tools), a GitHub Action, and D3.js visualization. 📊 Experimental Results ・Token efficiency: 38x-528x reduction (median ~82x across 6 repos) ・Impact prediction F1 score: 0.71 average ・CLI example: full context 12,921 tokens → graph context 762 tokens (~94% saved) #CodeReview# #AIAgents#

Cloudflare tried off-the-shelf AI code reviewers. We needed more control, so we built our own. It's now run over a million reviews. @_ryanskidmore helped make the build-vs-buy call and shares how teams should weigh control, cost, and maintenance in their decisions.