Skip to content
View laithalsaadoon's full-sized avatar

Highlights

  • Pro

Block or report laithalsaadoon

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
laithalsaadoon/README.md

Laith Al-Saadoon

Principal AI Engineer in AWS Developer Tools, Houston, Texas. I work on one idea, the autonomous work horizon: how long an AI agent can work on its own before a person has to step in. I run my own fleet of agents every day, measure that number, and try to make it longer.

My customers are the AI agents that build on AWS, and the people and companies behind them. I work with the field and with the companies that make coding assistants so whatever agent a customer already uses builds well on AWS, I turn what I learn into prescriptive guidance (instrumented open source, evals, and reference implementations that show one recommended path), and I carry what the field teaches me back to engineering.

Long-form writing lives at laithalsaadoon.com as the Frontier Field Notes.

Created

AWS MCP Servers stars

AWS MCP Servers, the open-source Model Context Protocol servers for AWS, now past tens of millions of PyPI downloads. I created the project and led AWS MCP strategy.

Building now

Repo What it is
microvms-agentd Exec and file-transfer daemon for AWS Lambda MicroVMs (Firecracker), with Rust, Python and Node bindings and a 75-check conformance suite. Agent sandboxes in your own account.
lambda-microvm-sandbox createLambdaSandbox() for the Vercel AI SDK harness, so the sandbox runs on Lambda MicroVMs under your network policy. Article.
erpaval Explore, Research, Plan, Act, Validate, Compound: the loop my coding agents run inside, with a lessons store that carries forward between sessions.

Earlier: the Generative AI CDK Constructs shown in Werner Vogels' re:Invent 2023 keynote, the PagerDuty and Amazon Q incident integration in Matt Garman's re:Invent 2024 keynote, and United Airlines' "Every Flight Has a Story" notifications on Amazon Bedrock and Claude. At AWS since 2016; before that one of the first ten at Onica (later Rackspace); before that a genomics lab.

Elsewhere

GitHub

GitHub stats for laithalsaadoon

Pinned Loading

  1. awslabs/mcp awslabs/mcp Public

    Open source MCP Servers for AWS

    Python 9.7k 1.7k

  2. sdk-python sdk-python Public

    Forked from strands-agents/harness-sdk

    A model-driven approach to building AI agents in just a few lines of code.

    Python

  3. awslabs/agent-plugins awslabs/agent-plugins Public

    Agent Plugins for AWS equip AI coding agents with the skills to help you architect, deploy, and operate on AWS.

    Python 890 151

  4. erpaval erpaval Public

    ERPAVal — autonomous software development. Six-phase Explore/Research/Plan/Act/Validate/Compound workflow with classifier-driven routing and a compounding lessons store.

    Python 28 2

  5. opencodehub opencodehub Public

    Apache-2.0 code intelligence for AI coding agents. Graph-based impact analysis, blast radius, and execution-flow tracing over MCP.

    TypeScript 3