ShipWithTez tests fast-moving AI tools in public, makes sense of what changed, and shows builders and teams which workflows are worth their time.
Built by Pranoy Tez. Engineering background across Microsoft and T-Mobile. Background
Builds show what I made. Workflows show how you can too, as take-home recipes with the exact prompts, commands, and artifacts. Copy what helps.
A take-home recipe that turns a bundled 50-thread sample inbox into a ranked action queue. Read-only pattern, human review required, zero credentials needed.
A take-home recipe that runs a 4-variant QA grid on a bundled sample prompt before any image lands on a public surface. Brand-safety, on-prompt fidelity, embedded-text check, alt-text candidate.
Paste your URL. The demo runs real checks against your homepage, robots.txt, sitemap.xml, and llms.txt, then hands you copy-ready prompts to verify whether each LLM channel can actually answer about your site.
The newest homepage-worthy build ships first so returning visitors see what is fresh. The rest of the shelf stays anchored to the strongest proof assets.
What held up
A useful stack critic has to start with evidence. This build turns one public URL into a sourced, confidence-labeled stack estimate before any scale advice.
Paste a website URL and get a public-evidence tech stack estimate from headers, HTML, scripts, cookies, and public web search.
What held up
A data-center atlas can do more than plot dots. Click a cluster and get a facility-class brief: MW scale, likely power architecture, cooling technology mix, regional constraints, and a reference blueprint before drilling deeper.
Click a major data-center cluster and understand what kind of facility it likely represents: AI mega-campus, hyperscale region, regional cloud cluster, or edge/enterprise footprint. Then open the reference blueprint or jump into the separate supply-chain map.
What held up
A useful company-research workflow does not need a terminal. A thin layer over SEC filing metadata and XBRL facts can make the recent filing changes legible in minutes.
Search a public company, inspect its recent 10-K, 10-Q, and 8-K filings, and compare what changed in the latest annual and quarterly filing facts.
What held up
Real-time public APIs, live animated ferry positions, highway camera feeds, and travel time congestion ratios across the Puget Sound region, built in one session.
A living portrait of the Seattle metro. Travel time corridors color-coded by congestion, ferries moving across the Sound in real time, live highway cameras on click, mountain pass conditions, and highway alerts.
The blog is where the build becomes legible: what changed, why it matters, and how the workflow actually worked.
Startups.RIP is a useful first example for AI in the Wild: failed startup history becomes product memory when agents can compare what was tried, why it broke, what changed, and what to test next.
Read →The interesting GPT-5.5 question is not whether it is smarter. It is where the model leaves less unfinished work behind, and whether that reduction is worth the premium routing cost.
Read →OpenAI shipped GPT Image 2 today. It is the new ceiling on text + UI mock realism. Google still owns cost. Black Forest Labs owns open weights. Ideogram owns dense typography. Firefly owns brand safety. Here is the honest map of the field after today, and a one-line recipe for picking the right model per job.
Read →9 years
Software engineering
27 builds
Shipped in public
3 AI providers
Tested per build
A short email when a new build, breakdown, or release note goes live. No spam, no schedule, just the useful parts.