DNdavidnguyen.io
02·/intelligence

Always-on market intelligence

A pipeline that watches your market every day and lands a written brief in the inbox, not a feed to scan, but a read on what moved and why it matters to the work in front of you.

Counts and channel types are public; the source list stays private by design.

captured run · daily pulsetriaging
crawled 1,8411,842kept 391signals 106107brief 1
queue23
voicesweb486items today34 sourcessig 4041
socialsocial641642items today31 sourcessig 28
newslettersemail38items today21 sourcessig 19
feedsrss676items today16 sourcessig 19
triage stream · sampled windowclass
Founder thread: agent-cost curves flattening06:41:07 · socialdropped low-signal
Changelog roundup: orchestration frameworks06:41:19 · feedsdropped duplicate
Launch promo: AI SDR tool discount week06:41:31 · socialdropped low-signal
Deep dive: eval harnesses hardening in production06:41:44 · newslettersauthored0.91
Job-board echo: GTM-engineer postings spike06:41:58 · feedsdropped off-taxonomy
Recap thread: same link, third occurrence today06:42:12 · socialclassifying…dropped duplicate
Pricing shift: usage-based tiers on agent platforms06:42:27 · voicesclassifying…authored0.88
Funding: enterprise-agent infra round announced06:42:41 · feedsclassifying…
today's run~100curated sources4channel types107surviving signals~6,600 wordsbrief lengthsampled window · sources private

Four channel types flow into parallel triage; surviving signals are mapped to active work and rendered into one brief.

How it works

It watches, so attention goes to judgment

A curated source graph is crawled, triaged down to real signal, mapped to active work, and delivered as a designed brief, automatically.

  1. 01

    Curate the sources

    Around a hundred hand-vetted sources across four channel types: professional voices, social, expert newsletters, and feeds. Counts and channels are public; the list is not.

  2. 02

    Triage in parallel

    Crawl results fan out to classification agents that drop noise, authored versus reposted versus mention, and tag every surviving signal against a fixed taxonomy.

  3. 03

    Map relevance, then deliver

    The differentiator: each signal is mapped to what's actively being worked on, with the so-what written out, then the brief renders to a designed PDF and lands in the inbox.

Proof

The engine, and a brief it produced

The intelligence pipeline as a repository, paired with a sample of the brief it lands in the inbox.

Engine

industry-pulse

A Claude Code pipeline that crawls LinkedIn, X, RSS, and email into one neutral daily signals brief.

Claude Code SkillsPythonMulti-source
What runs it
Claude classification agentsparallel triage + tagging
Multi-channel crawlersweb · social · email · feeds
Relevance mappingsignal mapped to active work
Automated PDF deliverydesigned brief to inbox

What your team gets

Stop reading the market by hand. A standing intelligence pipeline tuned to your context delivers a written, prioritized read every day, and the operator who can build the same for your team.