Backboard launches AI compression tool, coding assistant, and memory app from Ontario

·2 views

Canadian AI company Backboard announced four products on July 1, built around maximizing existing GPU efficiency rather than investing in new hardware. Its compression technology, BackboardQuant, reduces model size by up to 70% while maintaining full-precision performance and delivering up to 2.7x faster inference speeds. Backboard Studio, an agentic coding assistant, scored 79.8% on the Terminal-Bench 2.1 benchmark, outperforming Claude Opus 4.8's standalone result of 74.6%, and can run entirely on open-source models. The company also launched Nash, a consumer and enterprise chat app offering access to thousands of AI models with on-premise memory storage, which ranked first on two independent AI memory benchmarks. The entire stack is designed to run within a customer's own cloud environment, keeping data on-premises — a key requirement for sectors like healthcare, finance, and government.

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)

Developer Compares Free Online Table Makers, Launches Own Open-Source Tool

A developer evaluated several popular free table-building tools and found each lacking in key areas. Tools like Tables Generator and Markdown Tables Generator are functional but limited in scope, while Google Sheets and Notion are workarounds rather than dedicated table builders. A common shortcoming across all options was the absence of clean multi-format export capabilities. In response, the developer built Tablesmit, a free, open-source table builder that supports drag-to-resize, cell merging, and column types. Tablesmit allows export to PDF, Excel, LaTeX, CSV, and PNG without requiring an account.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer builds open-source manga reader with 38,000+ titles and privacy-first design

A developer has launched Mani Reader, an open-source manga reading platform built with Next.js 14, TypeScript, Tailwind CSS, and GSAP animations. The platform provides access to over 38,000 manga, manhwa, and manhua titles by leveraging the Consumet ecosystem, eliminating the need to manually maintain a content library. The project was motivated by frustration with existing manga sites that suffer from heavy ads, slow performance, and excessive user tracking. To address privacy concerns, Mani Reader includes an Incognito Mode that limits data collection without compromising usability. The codebase is publicly available on GitHub, and the developer is inviting contributions from the open-source community.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer builds Mermaid diagram editor, pivots to solving documentation drift

A developer built MermaidCreator, a Mermaid diagram editor featuring a visual canvas, code editor, and AI-powered diagram generation from prompts or whiteboard photos, using Next.js, Supabase, Stripe, and Claude. The core technical lesson was maintaining a single source of truth in the app's store, where visual edits, code edits, and previews stay in sync to avoid bugs. After completing the editor, the developer identified a deeper problem: architecture diagrams become outdated within weeks as codebases evolve, rendering them unreliable. To address this, the app was extended to index docs, Mermaid blocks, API routes, and dependencies from GitHub, Notion, Confluence, and Google Docs, flagging changes that affect existing diagrams for human review. Accepted fixes can be automatically pushed back as GitHub pull requests, though the developer remains uncertain whether the project will become a commercial product.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

AI Writes Code Fast but Cannot Own Production Systems, Engineers Warn

AI coding tools like Claude Code and Lovable can generate working prototypes in minutes, but engineers warn they fall short once software hits real-world production environments. IBM researcher Andy Anderson spent four months building a Kubernetes dashboard with AI assistance, only to lose more time reverting broken builds and fixing cascading errors than he would have spent coding manually. A large-scale study of over 302,000 AI-authored commits across nearly 6,300 GitHub repositories found that more than 15% introduced at least one issue, with 22.7% of those problems persisting long-term as silent technical debt. AI tools lack the contextual understanding needed to anticipate traffic spikes, dependency failures, or system-specific constraints that define production readiness. Engineers conclude that while AI accelerates prototyping, human ownership and judgment remain essential for building systems that survive reality.

0 comments Read more at DEV Community