Transparency, by default

Your voice. Your Life.
Your data.

Talvy only works if you trust it. It was intentionally built to protect your most personal thoughts, respect your privacy, and stay honest about the environmental cost of the technology we use.

Built for things you'd only tell a journal.

Your voice memos are personal. Talvy is designed with that in mind at every level: from where your data lives, to what we never collect, to how we handle the moments when processing does happen in the cloud.

Stays on your device
  • Your complete journal library
  • All memo cards and daily digests
  • Person profiles and connections
  • Themes, settings, and preferences
  • Browsing and search history within the app
Briefly touches the cloud
  • Raw audio while transcription is in progress
  • Memo text while AI parsing runs
  • Deleted from our servers immediately after processing
  • Never stored, indexed, or used for training

Encrypted in transit

Every request between Talvy and our processing services travels over TLS 1.3. Audio and memo text are never sent in plaintext. We don't log the content of your recordings, only the metadata needed to route and return the transcription result.

No ads. No trackers.

Talvy doesn't run ads and never will. We don't embed analytics SDKs, ad networks, or consent management platforms. No Google Analytics, no Meta Pixel, no Segment. Your listening habits and private thoughts are never packaged for anyone else.

We don't sell your data

Talvy makes money from subscriptions, not from monetizing your information. We don't sell, license, or share your data with third parties for commercial purposes. Your journal entries are yours, and that's exactly where they stay.

Not used for AI training

Your voice recordings and journal content are never used to train AI models, ours or anyone else's. The AI that powers Talvy's transcription and parsing is applied to your data ephemerally, then that data is gone.

Lightweight on your mind. Light on the planet, too.

Every app you use draws energy and water through the data centers behind it. We think you should know where Talvy stands, so we measured it. The numbers are small. We want to keep them that way.

Estimated monthly server-side electricity per user (kWh)
Talvy
~0.015
WhatsApp
~0.2
Gmail
~0.3
Spotify
~0.5
ChatGPT
~0.5
Instagram
~1.5
TikTok
~2.5
YouTube
~3.0
Netflix
~4.0

Talvy (light use, ~30 memos/month): ~0.015 kWh. Heavy use (~100 memos + daily digests): ~0.05 kWh. Figures for other apps derived from company-level energy reports divided by monthly active users: Google Environmental Report 2024, Meta Sustainability Report 2023. ChatGPT estimated from Li et al. (2023) at ~50 queries/month. Spotify scaled from audio-vs-video intensity ratios. Netflix from Obringer et al. (2021). See also: IEA Data Centres (2024).

~0.04 L
One month of Talvy
(light use, ~30 memos)
6 L
One toilet flush
(EPA WaterSense)
140 L
One cup of coffee
(Water Footprint Network)

Water figures use a conversion of 2.5 L per kWh, based on the US average water intensity for electricity generation (2.18 L/kWh, USGS) plus direct data center cooling overhead, per Mytton (2021), npj Clean Water. A single Talvy memo uses roughly the same water as one Google search.

A journal you can
trust completely.

Talvy is launching soon. Join the waitlist for early access.

... people already waiting
Get early access

No spam. Unsubscribe any time.

You're in. ✦

You're #... on the list.
Share your link to move up the waitlist.

talvy.app/early?ref=…