Silicon intelligence
Where does capability live: inside components, in their relations, or in the history of their interaction?
Explore intelligence →Scale output,
not headcount.
Growth usually demands more people, more specialist attention and more operating cost. EXUP looks for the opposite: more completed outcomes from the people and resources you already have.
EXUP turns scarce human attention into more capacity — and gets paid only from value we can verify.
Where is human attention the bottleneck?
Specialist review, coordination, analysis and exception handling often limit how much the organization can deliver.
Can the same team deliver more?
We look for measurable gains in completed outcomes per person, hour and unit of resource — without weakening quality.
Who carries the risk?
We do. If the agreed improvement cannot be verified, there is no value fee.
See how EXUP works →RESEARCH — SYSTEMS — REAL-WORLD OUTCOMES
THE CHALLENGE
Human attention limits capacity.
- More people as the default solution
- High specialist dependency
- Fragmented tools and processes
- Rising operating cost
- Unclear control and accountability
THE EXUP APPROACH
Turn repeatable work into verified capacity.
- Reuse proven workflows
- Let machines handle structured work
- Keep human authority where it matters
- Verify actions and accepted outcomes
- Measure completed output
OUTCOME
More output from the same team.
EXUP looks for measurable increases in accepted output without increasing the constrained human resource.
Illustrative model · actual value requires a measured baseline and verified result.Buy the outcome.
Inspect the machinery.
EXUP is the commercial front door. Machine Action Infrastructure is the technology behind governed economic or physical consequence. The broader product and research universe stays available without making a buyer decode it first.
Four connected lines
of inquiry.
Each starts with a question. Each needs a system that can be tested.
Execution economics
What is the lowest-cost valid way to deliver an accepted outcome, without losing quality or violating constraints?
Explore EXUP →Physical systems
What survives when intelligence meets hardware, infrastructure, materials and local conditions?
Explore physical intelligence →Human control
Can meaningful control be observed at the point of consequence, rather than merely declared in a policy?
Explore human control →
EXUP · CAPACITY
More accepted work.
Get more done with the people and resources I already have.
Human attention is the constraint. EXUP measures whether the same organization can deliver more accepted outcomes per person, specialist hour or constrained resource — and counts only verified value.
See the economics and proof →
PHYSICAL · REAL WORLD
Let the machine
do the job.
Operate this machine within its real operating conditions.
Perception, local intelligence and controlled physical action become one continuous job. The machine does not stop at a recommendation; it carries bounded work into the physical world and observes what happened.
Open VALO Physical →
CAI / PRE-NEED · LIVING STATE
Keep the problem alive.
Watch this, notice what changes, and move it forward before I have to reconstruct it.
The system observes signals, maintains state across time, detects material change and drives the next appropriate step. PRE-NEED moves the loop earlier: from waiting for a request to recognizing when a need is emerging.
Explore CAI →
RELATION · CONTINUITY
Handle the relationship.
Remember what matters between us and handle what follows.
relAIon is not a chatbot session. The job is continuity: shared history, commitments, change and the next relevant action survive across time and replaceable models.
Product direction · bounded claims only
Inspect what exists.
Executable demonstrations, public evidence and explicitly labelled concepts. A representation is not proof of an outcome.
Evidence case
Cost & capacity
Inspect the published execution-economics case and its claim boundary.
Open evidence →Interactive demonstrationAuthority that changes
See what happens when permission changes before an action takes effect.
Run the demo →Physical platform · conceptHabitat
Explore the engineering, logistics and economic assumptions behind a composable habitat.
Explore the concept →UNDER THE JOBS
Inspect the machinery.
The complete technical and research map remains here for people who need to inspect how the jobs are built. It is infrastructure and research depth, not the buying story.
View the complete product and research map
Machine Action InfrastructureAction boundaryVALO sits between machine intent and real-world consequence: Model → Intent → VALO → Action → ProofOpen →
VALO PhysicalPhysical AI · Edge infrastructureLocal models, sensing and governed machine action — from edge inference to physical consequence and evidenceOpen →
HEIMELOpen authority-to-consequence contract · v0.2.0Public contract, formal models, 79 normative conformance vectors, reference implementation and signed release evidence. HEIMEL exposes the public consequence-time authority boundary; the broader VALO / REHT implementation remains separate.Release →
MAInteraction architecture · conceptMaking the real world digitally understandable, addressable and interactable across devices, interfaces and physical systemsOpen →
CAICore · prototypePersistent intelligence missions: discover, detect change, evaluate, act with authorityOpen →
TraXinGCUCoreEconomic and operational work unitWork contract → settlement
Factory OSPlatformProduction runtimeCanonical runtime
Factory LineProductSellable production line over Factory OS / GCUOutcome delivery
AI-native IPShared engineCanonical entity, identity, rights and provenance across OLAV, AI Management and microdrama.Explore →relAIon / relAILayerCommunication and relational layerRelational systems
Just YouHuman centrePersistent person: identity, relationships, permissions, data, preferences and contextSystem centre
AI Captain’s BridgeAgent operations · buildingPersonal control plane for models, agents, tools, memory, channels and services around Just YouOperational layer
Personal News AgentJust YouPersonal change radar with source evidence and governed access · conceptOpen →
JustITSurfacePersistent address / execution surfaceExecution
Protocol LayerProtocolInteroperable protocol layerInteroperability
Edge / MCU / NPUEdgeLocal execution and physical consequence layerLocal runtime
Capability LayerLayerReusable capability product layerCapabilities
EmplAIWorkWork / capability product trackMachine work
Cleanroom / Model & Reasoning ProfilerProfilerModel and reasoning profilingEvaluation
Personal Sovereignty ConformanceConformanceCross-cutting conformance layerConformance
EXUP
CARE
Dyade
Roomit
GOI
VALO Habitat
BIDRAConcept · pilotIdeelle organisasjoner: penger + relevante handlinger + dokumentert effektOpen →Human Control Barometer (HCB)Experimental · retrospectiveIndependent measurement of human review, intervention and control. Human Control Research / Nishita track.Open →
Human Control ResearchResearchBehavioral evidence, bounded inference and falsifiable tests of meaningful human control in AI-assisted decisionsOpen →
VALOUmbrellaResearch / commercial umbrella and product surfaceVALO ResearchHow the systems connect
INTENT → CONSEQUENCEIllustrative network and machine scene · not experimental evidence.
Everything starts with you and returns to you.
Just You is the persistent human centre. AI Captain’s Bridge connects models, agents, tools, memory and channels around that person. CAI forms candidate action. HEIMEL decides whether it may happen now; REHT binds that decision to the exact consequence. EXUP finds the best valid execution path. Evidence returns what actually happened to Just You.
01
Just You
The persistent human object at the centre: identity, relationships, permissions, data, preferences and context. The surrounding system exists to serve the person, not to make the model the principal.
Human centre02
AI Captain’s Bridge
The operational layer around Just You: orchestrates replaceable models, agents, tools, memory, channels and services without making any of them authoritative by themselves.
Building now03
CAI
Find and evaluate what should happen. Persistent intelligence missions discover signals, detect change and form candidate actions.
Explore ↗04
HEIMEL + REHT
Decide whether the exact action still has authority now, then bind that fresh decision to the consequence boundary. Earlier admission or approval is not enough.
Explore ↗05
EXUP
Find the best valid way to execute the accepted outcome. Optimize cost, latency and runtime only inside the constraints that still hold.
Explore ↗06
Evidence / replay
Bind decision, execution and observed result so the outcome can be inspected, replayed and returned to the human centre.
Inspect evidence ↗Build. Break. Measure. Transfer.
A claim becomes useful when someone else can inspect what happened. Unknown remains unknown.
1
Make it executable.
Working systems expose assumptions diagrams hide.
2
Try to falsify it.
Change state, remove components, replay edge cases and keep negative results.
3
Keep the evidence.
Separate what was observed from what remains inference.
4
Transfer only what survives.
Turn stable findings into systems, standards, products or new tests.
The people behind the work.
VALO brings together product architecture, operational adoption and independent research. Each person’s contribution and relationship is stated below.
Njål Gaute Solland
Founder · Product and governance architecture
Leads VALO’s research direction, product architecture and consequence-time authority model.
Triin Solland
Operations and human adoption
Pilot implementation, onboarding, workflow integration and operational adoption.
Elsa Sklavounou
Authority Instrumentation™
Authority, identity, delegation, revocation and intervention boundaries. Joint case work examines how independent authority evidence meets VALO’s consequence-time checks.
Margaret Stokes
Creator of Aurora-Lens
Governed evidence admission, provenance, contradiction preservation and lawful continuation. Aurora-Lens remains independently owned; Margaret has no current operational delivery obligation to VALO.
Explore the research ↗Charles R. Rupp
Adjudication and justified reliance
Independent review and contribution on institutional authority, EFA/MECHA and the Three-Plane Model.
Jasper van de Meent
Causal ordering and execution continuity
Research on causal ordering, Lamport-anchored evidence and execution continuity informs VALO’s causal-evidence work.
Zaid Hasan Khan
Reliability and repeatable case testing
Contributed runnable synthetic healthcare case work on defined state transitions, stress testing, repeatability and preserved evidence. Further independent scrutiny has been invited.
Follow the evidence, not the claim.
Open the demonstrations, publications, code and reports behind the systems.