Hacker News: erenkaradag

New comment by erenkaradag in "Show HN: AI health agent that nags you on WhatsApp instead of a dashboard"

erenkaradag — Fri, 16 Jan 2026 21:57:44 +0000

Since Vitalify is an agent and not just a static checklist, it adapts. A standard app sees a missed workout as a “failure.” Vitalify sees it as data. If you tell the agent “I’m feeling under the weather” or “I’m swamped at work,” it doesn’t guilt you. Instead, it updates that day’s plan (e.g., swapping a workout for “rest & recover,” or scaling things down).

New comment by erenkaradag in "Show HN: AI health agent that nags you on WhatsApp instead of a dashboard"

erenkaradag — Fri, 16 Jan 2026 21:45:07 +0000

Appreciate that. We don’t integrate with wearables yet. We focused first on building the execution layer (ingestion → actionable plan → calendar + nudges). Wearable ingestion (Apple Health / Google Fit) is on the roadmap so trends like sleep/steps/RHR can adapt the plan, but we wanted the “what should I do next?” part to be solid before syncing more data.

Show HN: AI health agent that nags you on WhatsApp instead of a dashboard

erenkaradag — Fri, 16 Jan 2026 21:16:59 +0000

Hi HN, we built Vitalify.ai -started months before both ChatGPT and Claude Health announcements- because managing health data is fragmented and overwhelming.

Here's a video of the product: https://www.youtube.com/watch?v=LLI_1xB22rc

You upload your lab results, add your conditions/medications, and Vitalify generates a personalized wellness plan with daily tasks in your calendar.

While the value proposition might seem similar to others, we took a different approach with our UX. Instead of just a chat interface, we built a structured system with actionable plans, calendar, gamified goals, streaks and etc...

Data is encrypted at rest and isolated per user. No data is used to train any models. You have full ownership to export or delete your records at any time.

If you're not comfortable with how big health AIs handles your data, or don't want to wait on their waitlist/country restrictions, our early access is open.

Happy to answer questions and feedback is super welcome!

Comments URL: https://news.ycombinator.com/item?id=46652329

Points: 5

# Comments: 4

New comment by erenkaradag in "TimeCapsuleLLM: LLM trained only on data from 1800-1875"

erenkaradag — Mon, 12 Jan 2026 21:49:56 +0000

The problem is that the 'genius' of Einstein wasn't just synthesizing existing data,but actively rejecting the axioms of that data. The 1875 corpus overwhelmingly 'proves' absolute time and the luminiferous aether. A model optimizing for the most probable continuation will converge on that consensus.

To get Relativity, the model needs to realize the training data isn't just incomplete, but fundamentally wrong. That requires abductive reasoning (the spark of genius) to jump out of the local minimum. Without that AGI-level spark, a 'pure knowledge pile' will just generate a very eloquent, mathematically rigorous defense of Newtonian physics.

New comment by erenkaradag in "You Don't Hate the American Healthcare System Enough [video]"

erenkaradag — Sat, 10 Jan 2026 07:44:46 +0000

The point about "being your own doctor" hits the hardest. In the current -money first- healthcare systems, you are essentially forced to become the project manager of your own survival.

New comment by erenkaradag in "Forecasting the 2026 AI Winner"

erenkaradag — Thu, 08 Jan 2026 18:28:28 +0000

The fact that Anthropic has the best model with the least amount of data is a compelling plot twist, but considering that the latest Gemini 3 Flash has taken the lead in SWE benchmarks, either the talent at Google is better than you'd think, or data really does matter a lot.