How the Gera 2030-Infrastructure Fits Together
The problem each one addresses
GeraNexus — identity and access
In 2030 users will have dozens of AI agents acting on their behalf across services. GeraNexus is our identity + authorisation layer: sign-in with Gera, delegated agent permissions, scoped tokens, and auditable agent actions. It is the spine that lets GeraJobs talk to GeraClinic on behalf of a user without full account merging.
GeraMind — shared AI reasoning
Every Gera product that reasons about a user, a request, or a market uses GeraMind's shared reasoning infrastructure: prompt templates, tool catalogues, evaluation sets, observability. Instead of 31 separate prompt engineering projects, one reasoning substrate serves all.
GeraVoice — real-time voice
Voice is the fastest-growing interface in 2026 and will be the default in many categories by 2030. GeraVoice provides real-time, low-latency voice I/O (GPT-4o-class and beyond), usable by any Gera product — from ordering on GeraEats to studying on GeraLearn.
GeraLens — vision and world model
Cameras are compute. GeraLens handles vision: product recognition for GeraMarket, property scans for GeraRent, skin-condition triage for GeraClinic, AR overlays for games. A single vision pipeline, many product integrations.
GeraWitness — verifiable provenance
As AI-generated content explodes, provenance matters. GeraWitness provides cryptographic attestations for: user-submitted media (was this photo taken, not generated?), AI-generated content (what model produced this, with what system prompt?), and transactions (what happened, when, signed by whom?). Useful in-product (GeraSure claims evidence, GeraJobs CV authenticity) and as a standalone service.
How they interlock
- GeraNexus signs-in the user → access to all products.
- GeraMind provides the reasoning substrate for every product feature.
- GeraVoice is the interface layer when the user speaks (default on mobile by 2028).
- GeraLens is the interface layer when the user captures or shares media.
- GeraWitness signs and logs everything non-trivial — so future disputes have evidence.
Why build them now (not in 2029)
Each platform is useful today with current-generation models, but is designed to benefit from 2028–2030 improvements without rearchitecture. Building the interfaces early means every new product we add inherits them for free.
How it relates to Gera Prime
The 2030-infrastructure platforms are the “plumbing” under Gera Prime. A Prime member gets better limits across every product because the shared infrastructure is optimised for them. See the Prime model.
Related reading
Why we built 31 products · Gera vs single-product incumbents · How AI agents find Gera
Explore each 2030 platform.