Experience and results

Stefano Salvucci

12 years writing software, designing architectures, and automating processes. I take a problem, analyze it, and ship a solution that works.

Co-founder of Social Academy and BusinessinCloud โ€” Italy's first all-in-one educational marketing platform โ€” I've built digital products that serve millions of users and handle multi-million euro transactions.

Selected in 2019 as a YTILI Fellow by the U.S. Department of State and the German Marshall Fund, I collaborated with European entrepreneurs and leaders on innovation projects. Author of awesome-italian-pa-opensource, a curated list of open-source libraries and tools for integrating Italian Public Administration services into software applications. I supported the launch of Dock Startup Lab at Roma Tre University. For 2 years I taught Full Stack Web Development at Epicode.

Registered member of the Rome Engineers' Association.

๐Ÿค–AI automation
๐Ÿ—๏ธScalable architectures
๐ŸŒInternational track record
๐ŸŽ“Technical training

Published articles

Deja-vu: Zero-Dep Memory Layer for Coding Agents
AI AutomationStefano SalvucciJul 16, 2026

Deja-vu: Zero-Dep Memory Layer for Coding Agents

GitHub repo vshulcz/deja-vu delivers a memory layer for agents like Claude Code and Codex, with search, MCP recall, auto-context and secret redaction.

QuantumByte: Open-Source Engine Turns Intent into Apps
AI AutomationStefano SalvucciJul 15, 2026

QuantumByte: Open-Source Engine Turns Intent into Apps

GitHub's QuantumByte is an open-source app builder that generates working apps from intent prompts, verifying each step against business requirements.

Ditto Mines Claude Logs into Local you.md Profiles
AI AutomationStefano SalvucciJul 14, 2026

Ditto Mines Claude Logs into Local you.md Profiles

GitHub tool that extracts personalized AI agent profiles from Claude Code and Codex logs for private, productive coding workflows.

LLMs: Love the Tech, Skip the Singularity Hype
AI AutomationStefano SalvucciJul 13, 2026

LLMs: Love the Tech, Skip the Singularity Hype

Hacker News post praises real LLM progress while calling out negative hype and exaggerated claims, offering a grounded view for working developers.

Sqlsure: Semantic Checks for AI-Generated SQL
AI AutomationStefano SalvucciJul 12, 2026

Sqlsure: Semantic Checks for AI-Generated SQL

Sqlsure adds deterministic semantic checks for AI-written SQL, catching issues like fan-out double-counting and policy breaches before queries run.

dnsglobe: Global DNS Propagation Checker TUI on World Map
Software EngineeringStefano SalvucciJul 11, 2026

dnsglobe: Global DNS Propagation Checker TUI on World Map

New open-source GitHub tool lets you watch DNS records propagate across 34 worldwide resolvers via a terminal TUI with interactive map.

OpenAI Releases GPT-5.6: Sol, Terra and Luna Variants
AI AutomationStefano SalvucciJul 10, 2026

OpenAI Releases GPT-5.6: Sol, Terra and Luna Variants

OpenAI launched GPT-5.6 family with Sol, Terra and Luna, promising major gains in coding efficiency and cybersecurity for Node.js and React API users.

AI Changes the Economics of Software Rewrites
AI AutomationStefano SalvucciJul 9, 2026

AI Changes the Economics of Software Rewrites

AI is slashing the cost and risk of legacy codebase rewrites by leveraging well-known patterns in popular stacks like Node.js and React.

GPT-5.6 Sol, Terra and Luna Launch Publicly Thursday
AI AutomationStefano SalvucciJul 8, 2026

GPT-5.6 Sol, Terra and Luna Launch Publicly Thursday

OpenAI to launch GPT-5.6 Sol with Terra and Luna this Thursday, expanding global preview access. New automation use cases for Node.js, Python and React devs.

Ternlight: 7MB Embedding Model Runs in Browser via WASM
AI AutomationStefano SalvucciJul 7, 2026

Ternlight: 7MB Embedding Model Runs in Browser via WASM

Ternlight is a 7MB embedding model that runs in-browser via WASM, enabling client-side semantic search and RAG for React/Next.js apps with zero backend and full privacy.

Study: Code Cleanliness Impacts AI Coding Agents
AI AutomationStefano SalvucciJul 6, 2026

Study: Code Cleanliness Impacts AI Coding Agents

New minimal-pair research shows cleaner codebases improve autonomous coding agents' performance over those with static analysis violations.

Mouse Brings Surgical Precision to AI Coding Agents
AI AutomationStefano SalvucciJul 5, 2026

Mouse Brings Surgical Precision to AI Coding Agents

New coordinate-based editing tool with atomic rollback and contextual guidance fixes core limitations of AI agents like Claude and GPT when editing Node.js, React and Python codebases.

Jamesob's GitHub Guide to Running SOTA LLMs Locally
AI AutomationStefano SalvucciJul 4, 2026

Jamesob's GitHub Guide to Running SOTA LLMs Locally

Jamesob shares a detailed GitHub guide for running state-of-the-art LLMs locally, covering hardware options, costs from $2k-$40k, and Docker setups.

HN Explores New LLM Approaches for Coding
AI AutomationStefano SalvucciJul 3, 2026

HN Explores New LLM Approaches for Coding

Hacker News users discuss alternatives to prompt-response loops with LLMs to improve coding flow and productivity.

Kimi K2.7 Code Now Generally Available in GitHub Copilot
AI AutomationStefano SalvucciJul 2, 2026

Kimi K2.7 Code Now Generally Available in GitHub Copilot

GitHub releases Kimi K2.7 Code, the first open-weight model in Copilot, giving developers more choice and lower-cost options for daily coding workflows.

Trump Lifts Restrictions on Anthropic Mythos and Fable Models
AI AutomationStefano SalvucciJul 1, 2026

Trump Lifts Restrictions on Anthropic Mythos and Fable Models

The US has removed export restrictions on Anthropic's Mythos and Fable models, restoring public access to the most advanced AI models after weeks of talks.

Godcoder: Local-First Open-Source Desktop Coding Agent
AI AutomationStefano SalvucciJun 30, 2026

Godcoder: Local-First Open-Source Desktop Coding Agent

New GitHub project Godcoder launches a desktop AI coding agent that stays local, uses your own LLM keys and builds its own harness for privacy and control.

Bash4LLM+ Lightweight Bash Wrapper for LLM APIs
AI AutomationStefano SalvucciJun 29, 2026

Bash4LLM+ Lightweight Bash Wrapper for LLM APIs

A dependency-free Bash script brings LLM APIs to the terminal, offering easy integration for Node.js, Python and Rails developers.

Wayfinder Router routes queries between local and hosted LLMs
AI AutomationStefano SalvucciJun 28, 2026

Wayfinder Router routes queries between local and hosted LLMs

New CLI tool offers deterministic routing of LLM queries between local and cloud models with no extra calls. Balances cost, latency and privacy.

Peerd: First Browser-Native AI Agent Extension
AI AutomationStefano SalvucciJun 27, 2026

Peerd: First Browser-Native AI Agent Extension

Trending GitHub project peerd runs AI agent loops in Chrome/Firefox, driving tabs and sandboxed compute (JS, WASM) peer-to-peer with no backend or telemetry.