Human judgment, amplified by agents.
An independent AI product studio building tools for personal context, human relationships, and real-world decisions.
AI is abundant. Context, trust, and judgment are not.
Getyak builds products that help people preserve what matters, inspect how a system reached a conclusion, and stay in control when an action has real consequences.
| Product | What it helps you do | Current signal |
|---|---|---|
| Talent Signal | Turn recruiter-controlled conversation evidence into reviewable relationship context and one useful next step. | Early product foundation |
| DayPage | Capture raw daily moments and compile them into a journal and a portable personal knowledge layer. | Active build |
| Solo Compass | Discover concrete, story-rich experiences through an AI-curated living map for solo travel. | TestFlight preparation |
| Telepace | Keep real customer voices inside product decisions with voice-native, agent-first research infrastructure. | Public beta |
| Apply Agent | Organize a high-quality job search while keeping browser execution and final submission under user control. | Public design specification |
- CCT is a local-first capture, search, and analysis layer for Claude Code conversations.
- Rumu explores what a living, human-annotated library for canonical books could feel like.
- Evidence before inference. Keep every consequential claim traceable to its source.
- Humans approve consequences. Models may propose; people authorize external or irreversible actions.
- Reversible by default. Prefer local state, previews, checkpoints, and recovery paths.
- Complete loops over broad surfaces. Prove one useful path end to end before adding more.
Start with a real user job, show the evidence, keep the change reviewable, and use synthetic or explicitly authorized data. AI assistance is welcome; the contributor remains responsible for the result.
Contribution guide · Security policy · Community standards