Telnyx Tool Lets Developers A/B Test AI Prompts and Collect User Votes at the Edge

·1 views

Telnyx has released an open-source example app that enables developers to compare two AI prompt variants side by side using its Edge Compute platform. A user submits a single task along with two different system prompts, and the app runs both through Telnyx AI Inference simultaneously. Results and vote tallies are stored in a Stateful Actor, eliminating the need for a separate database. Users can vote on which prompt response they prefer, and aggregate statistics are available via a dedicated API endpoint. The project is intended to help teams make data-driven decisions about prompt changes rather than relying on subjective judgment.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Study: AI Coding Tools Generate Drift Up to 22x Costlier to Fix Than Human Code

A controlled study by ReWeaver AI tested five AI coding tools — Cursor, Claude Code, Lovable, Figma Make, and VS Code with Copilot — using 42 identical prompts across eight production-readiness dimensions, comparing outputs against a human-authored baseline. Researchers introduced a metric called the Production Drift Ratio (PDR), which weights how often code issues occur by their estimated remediation cost, rather than counting occurrences alone. While AI tools produced only 3.4 times the human drift frequency in Security and Privacy, the PDR for that dimension was 22 times higher, meaning the fixes were far more expensive and time-consuming. Across all eight dimensions, AI-generated code was consistently more costly to remediate than human-written code, even in areas where humans produced drift more frequently. Statistical testing confirmed the PDR gap was highly significant, while frequency alone showed no significant difference — highlighting that raw occurrence counts can mask the true cost of AI code quality issues.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How to Build Explainable Image Moderation Using Multimodal AI Chat Models

Developers can moderate uploaded images for NSFW and violent content by combining a vision-capable chat model with a strict JSON schema, rather than relying on a dedicated moderation endpoint. The approach uses a policy prompt, schema validation, and a conservative fallback, with decisions stored as normalized statuses like allow, review, or block for auditability. Content categories such as nudity, graphic violence, and hate symbols should reflect each application's specific rules, since thresholds differ across platforms like medical forums, marketplaces, and social products. A key engineering lesson highlighted is cost management: one evaluation run consumed 18.7 million input tokens — about 3.4 times the estimate — largely due to duplicated policy text across retries and image variants. The recommended fix is to measure prompt tokens upfront, run small batches first, and track cost per accepted decision rather than cost per raw request.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer Seeks Contributors for Zentrail, an Open Source AI-Native Desktop IDE

A developer has announced Zentrail IDE, an open-source desktop integrated development environment designed to enable multiple AI agents to collaborate with developers within a single workspace. The project aims to support features such as multi-agent task orchestration, AI-assisted code reviews, workspace memory, and integration with tools like Claude Code, Gemini CLI, and local AI models. Zentrail is being built using a stack that includes TypeScript, React, Tauri v2, Go, Python, and LangGraph, among other technologies. The project is currently in its early architecture and planning phase, and the team is actively recruiting contributors across frontend, backend, AI, design, and documentation roles. No prior expertise is required, and the initiative is open to students, hobbyists, and experienced engineers alike.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

AWS Open-Sources Kiro Crew, an Autonomous Multi-Agent Engineering Platform

AWS launched Kiro Crew on Tuesday, an open-source orchestration platform designed to move software teams from interactive AI assistants toward fully autonomous engineering workflows. The system coordinates multiple AI agents across repositories and developer tools, maintaining project context across sessions and continuing work even when developers are offline. Before its public release, the platform existed internally at Amazon as MeshClaw and was adopted by more than 39,000 Amazon builders in under six months. Kiro Crew can be self-hosted on local machines or virtual machines, addressing enterprise security and compliance requirements through sandboxing, signed audit logs, and a monitoring dashboard. AWS also released reference applications such as DevFleets, Issue Radar, and Task Runner to demonstrate real-world use cases including pull-request triage, dependency upgrades, and long-running engineering tasks.

0 comments Read more at DEV Community