founder's note · why_this_exists.md

I wrote a diary for
twenty years.
Then I couldn't anymore.

DailyVox is a free voice journal for iPhone with on-device AI, built by Karthikeyan NG. I stopped writing — and started talking. That's how DailyVox was born, and why it runs entirely on your phone instead of someone else's server.

20years
of diary writing
1solo
engineer
0rpc
cloud calls
0trackers
on your device
§01

The problem.

context

I kept a diary for over 20 years. Almost every day, I sat down and wrote about my life. It wasn't discipline — it was just something I did. Like brushing my teeth.

Then, in recent years, I started missing days. One day missed. Then a week. Then I'd look back and realize I hadn't written anything in a month. The habit that had been part of my life for two decades was disappearing, and I could feel it happening but couldn't stop it.

I tried different apps. Different formats. Different routines. Nothing stuck. Writing every day had become a chore. And when something becomes a chore, you stop doing it.

§02

The moment.

insight

Then I tried speaking instead of writing. And everything changed.

When you speak, you don't worry about grammar. You don't edit yourself. You don't stare at a blank page wondering what to say. You just talk. About your day. About what's on your mind. It flows — naturally, honestly, quickly.

Two minutes of talking captures more than ten minutes of typing. And it feels nothing like a chore.

Twenty years of writing in diaries, and the solution to keeping the habit alive was to stop writing.

§03

Building DailyVox.

decision

The problem was: there was no good tool for voice journaling I could trust with my most private thoughts. Every app I found either sent my audio to cloud servers for processing, or charged a subscription, or both.

I'm a developer. So I built my own.

DailyVox is simple at its core: open the app, press record, talk about your day — and SFSpeechRecognizer transcribes everything on-device with requiresOnDeviceRecognition = true. No cloud. No servers. No account. No subscription. Your voice literally never leaves your phone.

AI tools made this possible as a solo developer. What would have taken a team and months a few years ago, I shipped from idea to the App Store in a fraction of the time. That's the power of this moment in technology — one person with the right tools can build something real.

§04

The Digital Twin vision.

north star

After I started building DailyVox, I realized it could go far beyond just recording voice entries.

If you talk to this app every day — about your thoughts, your decisions, how you react to things, what makes you happy, what frustrates you — it builds a picture. A detailed, honest picture of who you are.

That's when I built the Twin. A multi-dimensional model of communication style, emotional signature, a personal knowledge graph of people and topics, and temporal mood predictions. All derived on-device from NLTagger signals. All serialized as Codable JSON in a single Core Data entity. All yours.

I want the Twin to eventually act like you. Speak like you. Respond like you. A true digital version of yourself — private, on-device, yours.

We already leave digital traces everywhere — social media, messages, emails. But all of that is scattered across platforms owned by other companies. None of it is organized. None of it is truly yours.

What if you had one place where you deliberately stored the most honest version of yourself? Not the curated version you put on social media. The real version.

I believe keeping a digital version of yourself will be an important part of the near future. Not science fiction — practical. And DailyVox is built to make that happen.

But not by sending your data to a server. Everything works on-device. Your Twin lives on your phone. Nobody else has access to it. I won't compromise on that.
§05

Why iOS only.

platform

DailyVox is on iPhone and iPad only. There's no Android version — yet.

That's a deliberate choice. Apple gives me strong security guarantees — the Secure Enclave, the Neural Engine, the privacy frameworks. The whole promise of DailyVox is that your data never leaves your device. I need to be absolutely sure I can keep that promise before putting it on another platform.

I'd rather be on one platform and keep your data safe than be on two platforms and compromise.

Android will come — after iOS is solid, after the user base grows, and after I'm confident the same privacy guarantees can hold.

§06

What I believe.

principles
PRIVACY

Privacy is the architecture.

Not a feature. Not a toggle. The entire stack is built so your data physically cannot leave your device.

ACCESS

Mental health tools should be free.

A diary shouldn't have a paywall. Nobody should choose between their wallet and their mental health.

LOCAL-FIRST

AI should work for you, not on you.

AI that processes your thoughts should run on your device, under your control. Not on a server training someone else's model.

HONESTY

Speaking is more honest than writing.

When you write, you edit. When you speak, you're real. The best insights come from unfiltered expression.

§07

The roadmap.

build log
Shipped · v1.0

Voice journaling + on-device AI

Tech: SFSpeechRecognizer (requiresOnDeviceRecognition = true) for zero-network transcription. NLTagger for sentiment, NER, POS. NSPersistentCloudKitContainer wrapping SQLite. AVFoundation AAC 44.1 kHz capture. CryptoKit AES-256-GCM backups. LocalAuthentication for Face ID / Touch ID. WidgetKit + AppIntents. All inference on Neural Engine.

Shipped · v1.1

Twin + personality model

Tech: Custom TwinEngine. CommunicationStyle (TTR, expressiveness, directness, formality, signature words). EmotionalSignature (valence / arousal / dominance baselines, time-based patterns, trigger topic correlations). PersonalKnowledgeGraph via NER mapping entities with emotional weights. TwinPredictions for day-of-week mood forecasting. All models as Codable JSON in the Core Data AIState entity.

Shipped · v1.2

Ask Your Twin + social sharing

Tech: TwinChatView with pattern-matched query system. TwinResponseGenerator assembles natural-language replies from Twin data. ShareablePersonalityCardView renders cards via SwiftUI ImageRenderer at 3x (Instagram Stories 1080×1920, X 1200×675). UIActivityViewController for native share sheet. SKStoreReviewController at milestone entries with 90-day cooldown.

Shipped · v1.3 + v1.3.5

Constellation Update + Dynamic Island

Tech: SwiftUI Canvas-rendered constellation (every entry becomes a mood-coloured star, 60 fps). Celestial onboarding flow with star-birth welcome. Ivory theme + sage/terracotta/gold palette. SF Rounded typography. ActivityKit Live Activities for the Dynamic Island — recording timer, "new star appeared" celebration, opt-in streak indicator. Lock Screen Constellation widget via WidgetKit. iOS 18 dark + tinted app icon variants.

Shipped · v1.4.0 + v1.4.1

The warm redesign

Design: a complete visual rework to a warm, calm palette — sage, gold, and terracotta on ivory — across every screen, with the Digital Twin moved to the center of the app. SF Rounded typography, warmed light and dark themes, refined onboarding, and a polished Journal. v1.4.1 made onboarding voice-first: the final welcome step invites you to speak your first entry, and your own words become the first star in your constellation.

Shipped · v1.5.0

Body Twin — HealthKit

Tech: HealthKit reads sleep, HRV (SDNN), resting heart rate, steps, mindful minutes — all on-device, "Data Not Collected" label preserved. Per-entry HealthSnapshot in Core Data (~10 floats), each tagged with activity context detected via HKWorkoutSession + CMMotionActivity, so a workout isn't read as emotional stress. HR is stored as a delta from your personal hour-of-day baseline. Every captured signal waits in a review-and-discard queue before it reaches the Twin. The Apple Watch companion — wrist recording and heart rate sampled during the entry — is the next step in Body Twin, not part of this release.

Shipped · v1.6.0

Memory, measurable fidelity & the first ambient signals

Tech: semantic search over per-entry NLEmbedding sentence vectors (device-local index, meaning not keywords, with a measured abstention threshold so it declines rather than guesses). A first learned trait — openness — from a small on-device model validated on consented human writing, surfaced only past a data floor with an honest confidence. Per-prediction confidence bands. Voice biomarkers (pace/energy/pause/pitch → an "activated vs your own normal" arousal signal). Ambient Signals (folded in from v1.5.5): on-device Vision photo scene labels + music-library mood, through a review-and-discard queue — derived labels only, off by default. Plus a Spoken Words vocabulary so mis-heard names stick, and SpeechAnalyzer transcription on iOS 26. All measured against honest baselines; zero network calls.

Shipped · v1.7.0

Foundation Models + tool calling

Tech: Apple's on-device 3B Foundation Model via LanguageModelSession, grounded on the v1.6 memory index. The model never emits free prose — it emits a structured answer whose every factual sentence must cite a journal entry or a measured Twin signal, with citations constrained per turn so citing an unretrieved entry is impossible, and a deterministic audit checks every answer before it renders (a failed answer falls back to the classic chat). Read-only evidence tools via the Tool protocol; @Generable structured outputs; an adversarial evaluation battery (grounding, tone, false-premise and injection resistance) that the release passed twice consecutively. Requires an Apple Intelligence iPhone; the classic chat remains everywhere else. Zero network calls.

v1.8 · Multi-Language

The Twin speaks your language

Tech: Xcode String Catalogs for multi-language UI — Tamil, Kannada, Hindi, Spanish, Japanese, German first. Speech framework already supports 60+ transcription languages on-device; the UI catches up to the pipeline. iPhone-only by conviction — no macOS or visionOS targets. Depth on one platform, breadth in languages.

v2.0 · Apple Intelligence Native (iOS 27)

The Twin joins the system assistant — on your terms

Tech: Siri AI integration via App Intents — ask the system assistant and it consults your Twin. Cross-app context adoption as iOS 27 APIs open to third parties, with DailyVox's standing rule: context flows in, nothing flows out. Next-generation Apple Foundation Models for deeper reasoning and longer conversations. Twin replies in your voice via Apple Personal Voice.

v2.1 · Personality Depth

Persona conditioning — Twin learns to think like you

Tech: Validated Big Five personality scoring from your journal narratives, grounded in language-based personality modeling research. Multi-tier persona conditioning — the measured profile plus your actual phrasings, retrieved from the semantic index, steer every Twin reply. Monthly personality snapshots power identity-evolution diffs and "talk to your past self." Fully on-device; no desktop training step, nothing ever leaves the phone.

v3.0 · The vision

A true digital version of you

End state: an on-device AI that acts like you, speaks like you, responds like you — built from years of your own entries. Full RAG. Persona conditioning grounded in your measured personality. Tool calling for data retrieval. Session transcript with context condensation. Encrypted. On your device. Exportable only by you. Your data never touches a server. Your digital self is yours alone.

§08

Get in touch.

contact

I'm Karthikeyan. I built DailyVox because I needed it. Turns out other people need it too.

If you have questions, feedback, or just want to say hi —

install · $0 · no account

Start talking.

Free. Private. No account needed. Just open and talk.

Download free on the App Store