DEV Community • 2026-07-30 20:31

Relevance is not answerability: six signals, and none of them beat plain cosine

Part 2 of **The Answerability Problem. Part 1 showed the standard harness excluding the questions that test refusal, my own system scoring 0.000, and every lever I ship failing. It asked whether any cheap signal can separate "relevant" from "answerable." This is the bake-off. Code: RE-call. LongMemEval is the other public benchmark in this space, and unlike LOCOMO it names abstention as a firs...

0 0
10m read
Congrats to the DEV Weekend Challenge: Passion Edition Winners!
DEV Community • 2026-07-30 20:24

Congrats to the DEV Weekend Challenge: Passion Edition Winners!

We are excited to announce the winners of our DEV Weekend Challenge: Passion Edition! The prompt was simple and wide open: build something inspired by passion. You gave us rivalries, obsessions, fandoms, late-night side projects, and love letters to the things you cannot stop thinking about, and you built all of it inside a single weekend. Thank you to everyone who submitted. These projects were...

0 0
5m read
Hacker News • 2026-07-30 20:24

Show HN: Collie – a local AI harness that runs the browser, desktop and code

Comments

0 0
1m read
DEV Community • 2026-07-30 20:23

What Learning Without AI Taught the Last Pre-AI Generation

Originally published on robatdasorvi.com I learned to code in 2013 by being confused, searching for help, being confused again, and then eventually understanding something. The understanding was slow. It was uncomfortable. It was also, I now think, irreplaceable. My nephew is twelve. He has never written code without an AI suggesting the next line. He has never had to hold a problem in his head ...

0 0
8m read
DEV Community • 2026-07-30 20:23

Why Do Multi-Agent AI Systems Fail at Production Scale?

Originally published on robatdasorvi.com Last month I watched a customer support agent refuse to answer a question it had already answered three messages earlier. A second agent in the same pipeline had flagged the topic as sensitive. The first read the flag, deferred to it, and apologised. Neither agent was wrong by its own rules. Together they produced an outcome nobody intended. I keep thinki...

0 0
10m read
DEV Community • 2026-07-30 20:21

How to anonymise sensitive data in PDF documents with Python

Have you ever needed to share a bank statement with someone? Maybe you've needed a user of your project to share one with you for troubleshooting. I hit this exact problem. I've been building an extensible tool to convert UK bank statements into a relational database, but I only have access to statements from my own bank. How can I get access to other bank statements without asking people to send...

0 0
4m read
DEV Community • 2026-07-30 20:20

OpenAI Says Two API Settings Tripled GPT-5.6 Sol's ARC-AGI-3 Score

OpenAI says that two Responses API settings, retained reasoning and compaction, raised GPT-5.6 Sol's public-set ARC-AGI-3 score from 13.3% to 38.3%. The result matters because it shifts part of the explanation for a weak benchmark outcome away from the model alone and toward the software harness used to run it. OpenAI also reports that the configuration reduced output tokens by about six times. T...

0 0
5m read
DEV Community • 2026-07-30 20:19

A Year of AI Pair Programming: What Actually Changed

Originally published on robatdasorvi.com The first sign that something had shifted was when I stopped feeling proud of finishing things. That sounds worse than it is. I don't mean I stopped caring about the work. I mean the small dopamine hit of writing a function from scratch and watching it pass, that feeling got quieter. Not gone, but quieter. Replaced by something more like project managemen...

0 0
9m read
DEV Community • 2026-07-30 20:17

Cloud PBX for Legal Sector: Technical Architecture Guide

Cloud PBX gets pitched as a phone system upgrade. That framing undersells what it actually is: a real-time telephony application that connects VoIP infrastructure to business software through APIs and webhooks, and turns voice into structured data that other systems can act on. The legal sector is a good technical case study for what this actually looks like in production. Law firms have specific...

0 0
8m read
DEV Community • 2026-07-30 20:17

Choose the Reporting Grain Before You Choose the Chart

A chart can be perfectly formatted and still be wrong. In accounting and operational reporting, the most common cause is not the visual—it is the grain of the data underneath it. Grain means what one row represents. One row might be an invoice, an invoice line, a payment, an account balance at a reporting date, a CRM deal, or a single activity. Those are different business events. When they are ...

0 0
4m read
DEV Community • 2026-07-30 20:14

The AI-memory benchmark everyone quotes forbids saying “I don't know”

Part 1 of **The Answerability Problem. A follow-on from Retrieval-Augmented Self-Recall. That series argued agent memory needs calibrated abstention. This one is what happened when I actually tried to measure it. Code and every number: RE-call (MIT). There's a benchmark war going on in AI memory, and it's a healthy one. Mem0's paper (arXiv:2504.19413, Table 2) reports an LLM-as-judge score of...

0 0
8m read
DEV Community • 2026-07-30 20:11

Structuring Go Projects For Growth with Packages

Previously, we built a complete CRUD API for managing tasks. Although everything works, our entire application currently lives inside a single main.go file. This approach is fine for small examples, but as projects grow, a single file quickly becomes difficult to navigate and maintain. In this article, we'll reorganize our API into multiple packages, giving each part of the application a clear r...

0 0
7m read
DEV Community • 2026-07-30 20:11

Make Melodies and Chords with AI on this website

I'm a dev who's been producing electronic music for about 12 years, and the two halves of my brain finally collided into something. The problem: producers lose enormous amounts of time to the blank piano roll. Not to mixing, not to sound design — to the moment where you have the vibe in your head and no idea what notes to click. Existing tools either generate a whole finished track (useless — nob...

0 0
1m read
GenRec: Towards LLM-Native Recommendation at Netflix
Netflix TechBlog - Medium • 2026-07-30 20:10

GenRec: Towards LLM-Native Recommendation at Netflix

Authors: Ying Li, Arjun Rao, Shradha SehgalIntroductionRecommendations sit at the heart of the Netflix experience. Our current production models rely on thousands of hand‑crafted features over users, items, and interactions, along with specialized architectures for sequence modeling, feature interactions, and multi‑task objectives. This stack has evolved over many years to support diverse content ...

0 0
14m read
DEV Community • 2026-07-30 20:03

Securing Azure Functions: Use Managed Identities Over Connection Strings

TL;DR: The Key Takeaways at a Glance The Problem: By default, Azure Functions use connection strings for storage accounts, granting full access if leaked. The Half-Measure: Azure Key Vault hides the string but doesn't fix the lack of granular permissions (full access remains full access). The Best Practice: Use Managed Identities combined with Role-Based Access Control (RBAC). The Implement...

0 0
4m read
DEV Community • 2026-07-30 20:03

He creado un bot de Telegram para gestionar los temas(topics) de los grupos

Hola, he creado un bot para gestionar los temas de los grupos de Telegram. Aunque sea algo relativamente nuevo (fue en 2022) no veo mucha gestión particular del asunto. He empleado TypeScript con grammy tirando de servicios gratuitos (SupaBase, Vercel) Y apoyándome en la IA para las dudas porque soy realmente programador Java y esto es un lío. Sus principales funcionalidades: El bot se resume en...

0 0
1m read
DEV Community • 2026-07-30 20:01

Can we play a game on pixelden? Made my pixel art game portal playable inside Claude via MCP

I hacked the new MCP UI extension 😎 By which I mean I read the spec and used it exactly as documented, for a dumb little game. 😅 MCP Apps is the official MCP UI extension. Short version: your server declares an HTML resource under ui://, links it to a tool via _meta.ui.resourceUri, and the host renders that HTML in a sandboxed iframe right in the conversation. The iframe talks back over JSON-RPC...

0 0
1m read
DEV Community • 2026-07-30 20:01

Build a Task Queue System with Python and Redis

Build a Task Queue System with Python and Redis tags: python, redis, architecture, tutorial tags: python, redis, architecture, tutorial tags: python, bot, automation, tutorial tags: python, security, tutorial, tools tags: python, security, tutorial, tools tags: python, security, tutorial, tools Imagine typing your master key once and instantly unlocking passwords...

0 0
6m read
DEV Community • 2026-07-30 20:00

Compose AI agents like reusable components

Imagine composing AI assistants the way you create reusable React components. One agent reviews correctness. Another plans the tests. Their outputs flow into a third agent that produces one release recommendation. Correctness ──┐ ├──> Synthesis ──> Release recommendation Test plan ────┘ That is agent composition: small specialists, explicit dataflow, one final result....

0 0
3m read
Product Hunt — The best new products, every day • 2026-07-30 20:00

DepthData

The system of record for your company's AI spend. Discussion | Link

0 0
1m read
Previous Next