{
  "version": "https://jsonfeed.org/version/1.1",
  "title": "AI Analytics — Technical writing",
  "home_page_url": "https://ai-analytics.org/writing/",
  "feed_url": "https://ai-analytics.org/feed.json",
  "description": "Long-form technical notes from building intelligence infrastructure.",
  "language": "en-US",
  "icon": "https://ai-analytics.org/logo-512.png",
  "favicon": "https://ai-analytics.org/logo-32.png",
  "authors": [
    {
      "name": "AI Analytics",
      "url": "https://ai-analytics.org"
    }
  ],
  "items": [
    {
      "id": "https://ai-analytics.org/writing/distributed-vpn-routing/",
      "url": "https://ai-analytics.org/writing/distributed-vpn-routing/",
      "title": "Building a distributed VPN with intelligent routing",
      "content_text": "How we route around censorship with ML-driven path selection, traffic morphing, and 142 entry-node IPs.",
      "summary": "How we route around censorship with ML-driven path selection, traffic morphing, and 142 entry-node IPs.",
      "date_published": "2024-10-15T00:00:00.000Z",
      "tags": [
        "Censorship",
        "VPN",
        "ML routing",
        "DPI evasion"
      ],
      "authors": [
        {
          "name": "AI Analytics"
        }
      ]
    },
    {
      "id": "https://ai-analytics.org/writing/nlp-pipeline-scale/",
      "url": "https://ai-analytics.org/writing/nlp-pipeline-scale/",
      "title": "NLP pipeline for real-time sentiment analysis at scale",
      "content_text": "Architecture of a real-time NLP pipeline: TensorFlow models, sub-2-second latency, multi-language sentiment + entity recognition.",
      "summary": "Architecture of a real-time NLP pipeline: TensorFlow models, sub-2-second latency, multi-language sentiment + entity recognition.",
      "date_published": "2024-09-28T00:00:00.000Z",
      "tags": [
        "NLP",
        "TensorFlow",
        "Infrastructure",
        "OSINT"
      ],
      "authors": [
        {
          "name": "AI Analytics"
        }
      ]
    },
    {
      "id": "https://ai-analytics.org/writing/election-anomaly-detection/",
      "url": "https://ai-analytics.org/writing/election-anomaly-detection/",
      "title": "Detecting election anomalies using statistical methods",
      "content_text": "Benford’s Law, turnout modeling, ARIMA time-series — surfacing anomalies worth a second look.",
      "summary": "Benford’s Law, turnout modeling, ARIMA time-series — surfacing anomalies worth a second look.",
      "date_published": "2024-09-12T00:00:00.000Z",
      "tags": [
        "Elections",
        "Statistics",
        "Benford",
        "OSINT"
      ],
      "authors": [
        {
          "name": "AI Analytics"
        }
      ]
    },
    {
      "id": "https://ai-analytics.org/writing/processing-millions-posts/",
      "url": "https://ai-analytics.org/writing/processing-millions-posts/",
      "title": "How we process 2.4M social-media posts per hour",
      "content_text": "Apache Kafka, TimescaleDB, 80 GPU NLP workers, MinHash deduplication. The pipeline behind 58M posts/day.",
      "summary": "Apache Kafka, TimescaleDB, 80 GPU NLP workers, MinHash deduplication. The pipeline behind 58M posts/day.",
      "date_published": "2024-08-30T00:00:00.000Z",
      "tags": [
        "Kafka",
        "TimescaleDB",
        "NLP",
        "Infrastructure"
      ],
      "authors": [
        {
          "name": "AI Analytics"
        }
      ]
    }
  ]
}