About Ttek2
Ttek2 is a topic hub and search engine for technology news. The product is the topic page — a single place where what's happening with a thing (a chip, a game, a release, an outage) is summarized, scored for momentum, and linked back to the publishers who broke it. Search is the second half: find what's been written across the open tech web on whatever you're curious about, with snippets, prices, and forum context surfaced inline.
We don't host original editorial. The opinions and reporting on this site live on the sites of the publishers we link to.
What's here
- /topics — more than 1,100 addressable topic desks. The current view emphasizes desks with fresh, source-backed activity; the broader library remains searchable. Active signals are refreshed throughout the day, normally on a 30-minute cycle.
- /search — search across articles, vendor stores, forum discussions, and topic hubs. Local full-text retrieval handles every query. When enabled and available, GPT-5.6 Luna may rerank a small set of eligible results; search falls back to the deterministic local ranking if that step is unavailable.
- Price tracking — the search SERP and topic pages show current prices for tracked Steam/GOG SKUs. "30d Low" badges appear when a product is at its recent floor. History is retained so longer-term trends become visible as data accumulates.
How signals flow
A crawler walks allowlisted tech publishers plus public RSS feeds and forum APIs (HN and public Reddit listings) several times per day. Articles are classified for tech relevance, deduplicated, and stored in a local SQLite/FTS5 index. A separate trending pipeline scores topics by mention frequency, source diversity, and momentum, then the AI Pulse layer (OpenAI GPT-5.4 Nano, with a GPT-5.4 Mini fallback) generates per-topic briefs, sentiment, and discussion themes. Topic and homepage digests are generated in background jobs and served from cache. An eligible uncached search may use the optional reranker described above; no AI is required to return search results.
We respect robots.txt, honor takedown requests, and don't bypass paywalls or scrape behind auth. The user-agent we crawl as is identifiable and reachable.
What this site does NOT do
- No third-party analytics or advertising trackers. We use narrow, cookie-free first-party measurements for page readership, search quality, and result usefulness. See the Privacy Policy for exactly what is recorded and how long it is kept.
- No tracking pixels or third-party JavaScript. Open the dev tools — there are no
<script>tags from other domains. - No advertising network. No AdSense, no Taboola, no programmatic ads.
- No shopping browser behavior. Prices surface contextually; we don't optimize for purchase intent or run affiliate retargeting.
- No newsletter (yet — if we add one, it will be opt-in).
- No paywall.
What's open about this
The codebase is a custom PHP application. Architecturally interesting bits — the topic taxonomy, the trending merge logic, the rerank pipeline, the stream-write digests — are documented in commit messages and inline comments.
If you're a publisher and want your content removed from the index, email us. If you're a researcher who wants to study how topic clustering shifts over time, email us — we may be able to share aggregate snapshots.
Funding
This site is funded out of pocket. There's a donation link in the footer for anyone who wants to help cover hosting and LLM API costs, including the background AI Pulse and topic-digest jobs. It's not required and never gates anything.
Contact
- General: contact@ttek2.com
- Takedown / publisher requests: same address — please include the URL(s)
- Security: same address, please use "security" in the subject; see Terms of Use for the disclosure policy
Changelog
Notable changes are posted when they materially affect what visitors see. Current work focuses on broader source coverage, better search-result quality and clearer freshness and provenance throughout the site.
Built and maintained by Mark Janssen.