huksley

/**
 * Based on https://nextjs.org/docs/pages/building-your-application/configuring/custom-server
 */

/* eslint-disable @typescript-eslint/no-require-imports */
/* eslint-disable no-console */
/* eslint-disable @typescript-eslint/no-empty-function */

const { createServer } = require("node:http");
const { parse } = require("node:url");

huksley

Local postgres

It is so easy to run locally, and you don't need a docker for that!

Installing & running

brew install postgresql
git clone https://gist.github.com/huksley/f74e422424c03d8e37e5dc04a20f64eb pg
cd pg

huksley

/* eslint-disable @typescript-eslint/no-require-imports */
const crypto = require("node:crypto");
const fs = require("node:fs");
const sharp = require("sharp");
const logger = console;
const Redis = require("ioredis");

const redis = new Redis(process.env.REDIS_URL || "redis://localhost:6379", {
  enableOfflineQueue: true,
  commandTimeout: 3000,

huksley

Agents have arrived and they need their own picks and shovels

We’re seeing a strong wave of AI agent applications increasingly capable of taking on and automating complex workflows. A lot of what is driving this is taking place in the model layer - with frontier models better able to plan and reason. As Sam Altman has outlined, the emergence of better reasoning capabilities in models such as OpenAI o1 provides the foundation for potentially quick advancement to highly capable AI agents that can understand context, plan and reason in order to make decisions, and then take actions in order to achieve goals.

However, improved models alone will not deliver us advanced and autonomous agents, and, as of today, most models remain poor planners and low-level reasoners. This means that most AI agents are still susceptible to errors that become compounded across multi-step processes, meaning they are unable to reliably take on complex end-to-end tasks. As a result, an ecosystem of enabling software and new architec

huksley

Responsible AI in the Enterprise

As enterprises seek to gain AI maturity, and move along the path from experimentation to scaling applications, greater efficiency and success can be achieved through the establishment of a holistic Responsible AI approach to identifying opportunities and managing risks.

BCG defines Responsible AI (RAI) as “an approach to deploying AI systems that is aligned with the company’s purpose and values while still delivering transformative business impact.”

Put so succinctly, it sounds simple, but in practice, it entails establishing the following, across a large and complex organisation:

• Strategy - Comprehensive AI strategy linked to the firm's values that ties back to risk strategy and ethical principles.

huksley

import type { ZodSchema } from "zod";
// eslint-disable-next-line no-duplicate-imports
import { z } from "zod";

// Define a Zod schema
const userSchema = z.object({
  // Fix 1. Do not use catch - wreks the input type https://github.com/colinhacks/zod/issues/2852
  // Fix 2. Use optional() before default(), don't use at all???
  page: z.number().default(1),
  sort: z.string().optional(),

huksley

const { GenericContainer } = require("testcontainers");

const container = new GenericContainer("ubuntu:latest")
  .withLogConsumer(consumer => {
    consumer.on("data", data => {
      console.log("LOG", data.toString());
    });
    consumer.on("err", data => {
      console.log("ERR", data.toString());
    });

huksley

/**
 * Robust implementation of request tracing. 
 * Just add getRequestTracker().withRequest(req) at the first instance of incoming http request and
 * it will persist request details during the whole request processing chain, async or not.
 * 
 * Similar how AWS X-Ray works, or how NextJS 13+ cookies() method were implemented.
 * 
 * Licensed under: MIT License
 *
 * Copyright (c) 2024 Ruslan Gainutdinov

huksley

Which of the following best describes your organization’s current usage or future intentions for using generative AI for each of the following specific use cases?

• Content categorization (e.g., auto-tagging and indexing media libraries, identifying medical or insurance codes, product descriptions, etc.)
• Fraud detection and prevention (e.g., detecting fraud in financial services, healthcare claims/billing, etc.)
• Chatbot/self-service for customer support (e.g., product help, searching documentation, etc.)
• Quality control and anomaly detection
• Customer feedback analysis
• Generating marketing/sales content (e.g., blogs, proposals, articles, etc.)
• Supply chain optimization and management (e.g., route optimization, loading schedules, etc.)

huksley

class Smoothmq < Formula
  desc "Real-time message queue with persisted topics"
  homepage "https://github.com/poundifdef/smoothmq"
  url "https://github.com/poundifdef/smoothmq.git",
      tag:      "v0.2.3",
      revision: "a023548ff08a8a8fab18b4e1011c81b56288a814"
  license "MIT"
  head "https://github.com/poundifdef/smoothmq.git", branch: "main"

  depends_on "go" => :build