Dev Systems - Building Systems That Scale

Show HN: Avoiding the Memory Wall by computing LLM inference directly inside RAM

The excitement surrounding PrismML’s 1-bit/ternary Bonsai models has the industry closely watching how smartphone giants, particularly Apple, will implement LLMs on edge devices. Moving AI on-device is a brilliant and necessary strategy. It ensures absolute user privacy in alignment with EU regulations, fundamentally shifts the economics away from costly cloud inference, and paves the way for a significant hardware upgrade supercycle as users seek true AI-capable silicon.To create a smart o

Tell HN: It's a Lot

I'm not sure if anyone needs to read this. I know I would, just to have some pause. A contemplative moment, if you will. Hence the post.--I need to know full-stack software engineering, preferably in full. That means going from hardware all the way up to whatever compiled language you choose. Web, app, desktop, all of it. Full-stack isn't just backend/front-end on web. I see a stack running way deeper and higher than that (this XKCD comes to mind [1]).But when I know all that, I a

Show HN: Running PrismML's Bonsai inside DRAM by breaking DDR4 timing rules

The excitement surrounding PrismML’s 1-bit/ternary Bonsai models has the industry closely watching how smartphone giants, particularly Apple, will implement LLMs on edge devices.Moving AI on-device is a brilliant and necessary strategy. It ensures absolute user privacy in alignment with EU regulations, fundamentally shifts the economics away from costly cloud inference, and paves the way for a significant hardware upgrade supercycle as users seek true AI-capable silicon.To create a smart on

Ask HN: How would you harden AI changes to a 1M-line legacy SaaS before review?

I’m not a software engineer, but I’ve been running an experiment to see whether agentic development could produce a useful prototype on top of an existing SaaS codebase.The codebase is 1M+ lines, 15 years old, hosted on Azure, and primarily written in C# and React.The prototype needs to be available for customer testing in September. No developers were available to work on it full-time, although I could occasionally get help with specific technical issues. An engineer will evaluate the implement

Why My Open-Source Project Hasn't Done Better

Since the beginning of 2026, many projects such as RTK, Caveman, and Ponytail have claimed that they can reduce token usage by 80–90%. Some of them gained tens of thousands of GitHub stars in a very short time.My previous startup was building an e-commerce chatbot. In fact, the product was almost identical to the Universal Commerce Protocol (UCP). After UCP was released, I decided not to continue pursuing that direction.Earlier this year, I started turning the architecture we had built for e-com

Show HN: Grounded-forge: RAG with summaries and task views precomputed at ingest

Hi everyone,I've been working on a knowledge retrieval system that can create distributable applications for several months and I'd love to share it with you!The issue I've found with asking LLMs for advice is that they are trained on a large corpus of text but don't have much discernment. It's even worse when you ask a question like "What is Agile?" Agile is an emergent property of strong leadership and a compatible organisation design, not a methodology to sh

Show HN: Syncle – Open-source real-time database synchronization engine

SYNCLE is a A system for synchronizing multiple databases in real-time to ensure consistent data across distributed services.

Show HN: A deterministic governance harness for agentic development loops

I genuinely think that autonomous code generation is the future. But today we are facing a problem: we cannot be confident enough in the result produced by an LLM. The problem is its way of working. Even with strong directive such as “make a button blue”, we cannot be sure that the button will be blue (maybe the context has contradictory informations). This POC is based on hooks and is validating that a deterministic generic harness is possible.For example, you may distribute a skill saying that

Reminder: Path to Production for Agents Webinar Series Starts Next Week

Next week, join Microsoft for the Path to Production for Agents webinar series—a free, six-session technical training designed to help organizations move from AI experimentation to secure, scalable, production-ready agent solutions. The series takes place July 27–28 and is aimed at architects, technical leaders, engineers, and AI practitioners looking to operationalize AI at enterprise scale.Across six expert-led sessions, you'll learn how to:Establish governance foundations for AI at scale

Distributed Systems Solve AI authenticity

As reliance on AI becomes mainstream, immutable cryptographic truth, through distributed ledger technology, emerges as a ...

Building a serverless AI assistant at Pelago: concept to care in two weeks

Healthcare organizations face a critical scaling challenge – how to maintain deeply personalized patient interactions as member bases grow, without overwhelming care teams or compromising quality. At Pelago, a digital health company specializing in substance use disorder support, the engineering team found a way to build an AI-powered solution to address this challenge using AWS services in just two weeks. In this post, you will learn how Pelago used AWS serverless and AI services, such as Amazo

AI Transport v0.6.0: mid-run steering

Steering lets a client redirect an agent while the agent is still working, so a follow-up message reshapes the answer in flight instead of cancelling the Run and starting over, or waiting for it to finish.In the AI Transport SDK, a Run encapsulates the agent's output for a single turn (including multiple iterations around an agentic loop). The last release, v0.5.0, made a single agent turn survive a crash by splitting a Run into re-attemptable Steps. This one is about what happens when the user

Building multi-Region resiliency for AWS CloudFormation custom resource deployment

AWS CloudFormation is the foundational tool of infrastructure-as-code for thousands of organizations running workloads on AWS. But as teams push the boundaries of what CloudFormation can do natively, custom resources have emerged as a powerful extension mechanism that unlocks a broad range of possibilities. Yet, when it comes to building resilient, multi-Region deployments with custom resources, customers quickly discover a gap: there is no built-in multi-Region support. In this post, we will ex

Architecting offline-first generative AI applications for edge deployments using AWS services

According to Siemens’ 2024 report The True Cost of Downtime, Fortune 500 companies lose an estimated $1.4 trillion annually because of unplanned downtime. This downtime is often worsened by a lack of skills to detect and resolve issues quickly. Generative AI offers a promising path to address this, but deploying these capabilities in industrial environments introduces a distinct architectural challenge: how do you bring the power of large-scale AI to locations where cloud connectivity is unrelia

Automate custom PII detection at scale with Amazon Macie and Step Functions

Organizations in regulated industries like financial services, insurance, healthcare, and government ingest large volumes of data containing personally identifiable information (PII). Your applications, claims processing systems, partner data feeds, and internal workflows produce files that may include names, addresses, Social Security numbers, and domain-specific identifiers such as policy numbers, member IDs, and medical record numbers. Identifying and classifying this sensitive data is a comp

Does Temporal Workflow Streams replace a durable session layer?

If you're running a customer-facing AI agent on Temporal, you've probably seen this: a user refreshes the tab mid-response. When the page comes back, the response resumes, but anything they were about to send back (an interruption, a tool approval) is gone. Temporal keeps the workflow itself crash-proof, but getting its output to a browser, and taking anything back from the browser into the workflow, has always meant building it yourself. The usual pattern is a Redis pub/sub relay (Redis ac

Show HN: Meltbox – where your agents send you briefs

Easy to try even without an account! Would love all feedback.Hey all, I'm building Meltbox as a better human in the loop system. It's a way to centralize all of the decisions you need to make into context-rich briefs in a system built for fast human review.[Why]Something major changed recently. Agents have gotten even better and even faster at building apps / sites / dashboards / etc... and that's changed my development workflow in a major way.I often have my agents

Coding Skills Development Report

BLOG: https://juntz-g1thub.github.io/#post=coding-skills-dev REPO: https://github.com/juntz-g1thub/CodingSkills.gitBackground Recently I've been experimenting with vibe coding. I'm using OpenCode as my development tool. When developing TUI programs, AI-assisted code modification proved extremely difficult — every small bug fix required repeated manual testing, and feeding those results back to the agent was cumbersome. Especially for TUI-level feedbac

Show HN: Scaffold a raw idea into something ship-ready in under a minute

Several months ago, I joined a WhatsApp community with around 500 people. Once a community becomes large enough, you start noticing certain patterns.Whenever someone announced a new product idea, the responses were usually the same:“That sounds amazing.”“I would definitely use it.”“Great idea.”“Can’t wait to try it.”I do not think people were intentionally being dishonest. Most were simply being kind. Nobody wants to be the person who discourages someone before they have even started.But that is

Introducing tfgen: configure your Terraform stacks using plain Go

Until recently, we extensively used HashiCorp's CDK for Terraform. Then they announced the end of its development. We didn't want to spend months migrating to a new ecosystem, but we needed a replacement.In the Infrastructure Team at Ably, we like Go. I wondered if there was an opportunity for something simpler.Provisioning infrastructure at AblyWe provision all our infrastructure for the core Pub/Sub platform where possible using Infrastructure as Code, initially using Terraform and then making