Sense what humans cannot naturally sense.
New physical, chemical, spectral or other environmental variables outside ordinary perception.
FRONTIER SCIENTIFIC COMPANY
Capabilities beyond the native human repertoire.
Aukemir is building the scientific and technological foundations to specify, engineer, test and ultimately integrate genuinely new human capabilities.
The objective is not a learning product. It is to make human capability itself an engineering frontier — from non-native perception and additional degrees of action to, where mechanism and evidence justify it, new functions supported by neural or biological substrate itself.
Current work is H1 / CP-01 Formal Capability Encoding, with Finite-Horizon Local Intervention Geometry (LIG) as the first controlled substrate for building the specification, measurement and falsification discipline.
H1 active CEM-E0 architecture only internal technical / synthetic evidence external executions returned 0 human evidence 0 H3–H5 evidence 0
00
What we are ultimately trying to make possible
FRONTIER AMBITION, EVIDENCE-GATED ENGINEERING
Make the human repertoire itself engineerable.
Humans evolved with a finite sensory, cognitive, motor and biological repertoire. Aukemir exists to test whether that repertoire can be deliberately expanded with functions that ordinary humans do not naturally possess.
Make non-native structure operational.
Functional access to computational or latent dimensions that ordinarily require explicit analysis rather than direct use.
Add controllable degrees of freedom.
Supernumerary effectors, remote embodiment or other forms of action beyond the native motor repertoire.
Gain functional access to hidden internal state.
Metabolic, endocrine, cardiovascular, immune or neural variables as possible synthetic interoceptive capabilities.
Where justified, let the organism itself support new function.
Biological / embodied de novo capability engineering is a central long-range destination class: new functions supported by neural or biological substrate itself, only after function-specific mechanism, alternatives, preclinical evidence, safety, controllability, ethics and governance justify the path.
These are program directions, not Aukemir achievements.Biological instantiation is a central long-range destination class, not a promise that every capability must reach H5. Current Aukemir evidence remains H1 technical/synthetic only.
01
Company / North Star
THE QUESTION BEHIND THE COMPANY
Can human functional repertoire itself become an object of engineering?
Humans can learn extraordinary skills. Aukemir is pursuing a different frontier: whether technology can eventually add integrated functions that are not part of the ordinary native repertoire.
Rnative repertoire
→specify · engineer · test · integrate
→R + Xnew integrated function
X is not a better dashboard, a memorized rule, restoration of a lost function, or an AI doing the task on a person's behalf. The long-term standard is a measurable added function whose acquisition and integration survive serious alternative explanations.
CROSS-SUBSTRATE METHOD HYPOTHESISCEM-E0 — ARCHITECTURE ONLY
specify X→independent truth→control leakage + confounds→strong alternatives→closed-loop use→transfer + integration→replicate→deepen modality only if justified
Aukemir has prospectively defined this as a capability-engineering hypothesis across distinct substrates. It has not yet demonstrated that the method generalizes beyond LIG.
Evidence boundary
No de novo human capability has yet been demonstrated by Aukemir. The North Star is a research and engineering direction, not a current result.
02
Technology
AN EVIDENCE-GATED LADDER
From formal substrate to deeper integration — without pretending the bridge is solved.
The H1–H5 ladder is a risk-reduction architecture. Each layer removes a different uncertainty. Progression is conditional, and not every capability must follow the same endpoint.
H1ACTIVE
Computational Capability Substrate
Specify a function, construct independent ground truth and make the substrate technically trustworthy.
H2NOT EXECUTED
Engineered Human Representation
Test whether a designed representation supports reusable human function rather than rule following, cue learning or better visualization.
H3FUTURE
Synthetic Capability Integration
Investigate genuinely non-native functional access through synthetic sensory, cognitive, motor, physiological or hybrid channels.
H4FUTURE
Neural Capability Integration
Explore neural access only if a demonstrated function has a measured bottleneck that safer external channels cannot adequately solve.
H5LONG-RANGE DESTINATION CLASS
Biological / Embodied De Novo Capability Engineering
Where a function-specific mechanism and evidence justify it, investigate whether the organismal substrate itself can support X.
The architecture is substrate-neutral. LIG does not become the future sensory signal, neural code or biological mechanism. Biological instantiation is mission-relevant but separately gated.
03
Current Program
CP-01 / FORMAL CAPABILITY ENCODING
Start with a function precise enough to falsify.
Every engineered capability needs a specification: what new function exists, how it is measured, what information it requires, how it is encoded, and how we distinguish integrated use from tool use. CP-01 is the first controlled place where Aukemir is building that discipline.
CP-01 is not the company endpoint. Its job is to build and stress-test engineering primitives that may later be reused on materially different capability substrates — only through prospective evidence, not by analogy.
01
Specify
Define the function and independent ground truth before designing the human-facing representation.
02
Encode
Transform relevant structure into a controlled representation without directly exposing the target answer.
03
Discriminate
Test against shortcuts, explicit instruction, strong visualization and machine delegation.
04
Transfer
Far transfer, novel action, reduced rule dependence, retention and replication are needed before stronger functional claims.
FIRST H1 SUBSTRATELIG
Finite-Horizon Local Intervention Geometry
LIG is a model-derived, target-independent representation of local finite-horizon response to a predefined bounded family of interventions. Full nonlinear simulation remains ground truth.
LIG is not the capability Aukemir ultimately wants to create. It is the first substrate where capability specification, independent ground truth, equivalence, leakage control, reproducibility and falsification are being made measurable.
It is not:
- global controllability
- structural controllability
- a classical controllability Gramian
- biological causality or efficacy
- the future biological mechanism
- proof of a new human capability
04
Evidence
WHAT IS TRUE NOW
Ambition does not promote evidence.
Aukemir separates architecture, technical construction, negative scientific results, external reproduction, human evidence, capability evidence and commercial evidence.
Current execution layerH1 / CP-01 activeCURRENT
Cross-substrate method stateCEM-E0 — architecture onlyUNPROVEN
Internal technical architectureBuilt / controlledBUILT
Internal synthetic / held-out technical evaluationCompleted internally in prior controlled stagesINTERNAL
Preserved internal scientific failures2 prior negative scientific attempts remain immutable in the current lineagePRESERVED
External computational executions returned0PENDING
Human participants / data0 / 0NOT BEGUN
Human learnability / far transfer / retentionNot demonstratedUNPROVEN
H3 synthetic capability evidence0 Aukemir evidenceFUTURE
H4 neural capability evidence0 Aukemir evidenceFUTURE
H5 biological capability evidence0 Aukemir evidenceFUTURE
Workflow / confirmed design partners / revenue0 / 0 / $0OPEN
Open Evidence Registry ↗
Internal technical success is not human evidence. Architecture is not method validation.
05
Build
BUILD RECORD
A sequence of questions, not a sequence of claims.
- 01
Scientific question
Define what would count as function rather than interface.
- 02
Falsifiable thesis
Predefine evidence that supports, weakens or kills the hypothesis.
- 03
Prior-art challenge
Search for closer precedents before claiming novelty.
- 04
Formal H1 specification
Define the first substrate precisely enough to reproduce and audit.
- 05
Calibration and metrology
Make the computational system trustworthy before asking a human question.
- 06
Representation architecture
Translate structure into a target-independent human-facing form with leakage controls.
- 07
Controlled experimental architecture
Separate transfer from memorization, explicit rules, visualization effects and machine delegation.
- 08
Synthetic integration
Exercise the end-to-end system before any human execution.
- 09
Negative scientific evidence preserved
Two internal scientific failures remain immutable; downstream work does not retroactively rescue them.
- 10
Strengthen H1 measurement controls
Current technical work is tightening prospective measurement and source controls before another scientific attempt.
- 11
Independent external reproduction
0 completed external executions returned in the current controlling record.
- 12
Human validation
Not begun. Requires separate authority, ethics and evidence gates.
06
Translation
COMMERCIAL WEDGE ≠ COMPANY IDENTITY
Useful intermediate technology can fund the frontier without becoming the frontier.
Aukemir is investigating complex biological-model workflows — including QSP, computational biology and systems biology — as an early commercial wedge.
The hypothesis is that model-heavy teams may value better intervention exploration, model criticism, fragility analysis and cross-disciplinary reasoning. That market is not validated.
SCIENTIFIC LEDGER
Capability evidence
Transfer, novel inference, integration, replication and biological instantiation remain separate scientific questions.
COMMERCIAL LEDGER
Workflow value
Problem severity, model access, design partners, measured pilot value and willingness to pay are tested separately.
Current commercial state
0 workflow evidence · 0 confirmed design partners · $0 revenue in the controlling records.
07
Founder
FOUNDER & RESEARCH LEAD
Isabella Salcedo Tuiran
Medical student, researcher and scientific founder working across biomedical science, complex systems and experimental human capability engineering.
At Aukemir, Isabella owns the scientific thesis, claim boundaries, capability architecture, current technical direction and company strategy. Future H3–H5 programs are explicitly expected to require specialist neuroscience, neuroengineering, biological, translational, safety and regulatory competence.
“
Treat human limits as hypotheses — and earn every step beyond them with evidence.
CONTACT
Criticism, collaboration and difficult questions are useful.
Aukemir welcomes rigorous scientific criticism, external computational reproduction, potential design-partner conversations, specialist collaboration and aligned frontier-company discussions.
[email protected] ↗