Frontier AI, independently evaluated.
A data-sovereign evaluation lab in boutique format — senior depth instead of crowd, sensitive material processed on-premises under NDA. Paired with CTC Research Lab: five workstreams on trustworthy evaluation and multi-agent safety, run on a single 32 GB GPU.
$ ctc eval run --rubric sig/rubric-v3.2.1#8f3a…e1 --lane batch RUN 2026-07-02T14:11Z · frozen rubric · hardened sandbox task 041/060 security_audit PASS grounded: repro ✓ · tests 12/12 task 042/060 code_generation PASS grounded: compile ✓ · lint ✓ · types ✓ task 043/060 code_understanding ESCALATE → arbiter · self-consistency σ 0.31 task 044/060 code_generation PASS grounded: compile ✓ · property tests ✓ … 60/60 · 4 min 57 s · ~200 tok/s · artifacts pinned #c41d…9a
Stylised render — run shape and numbers from the measured W2 pilot; output format illustrative.
Evaluation you can re-run.
Root-level evaluation of how frontier models reason, where they fail, and whether their behaviour holds under pressure — judged by a senior practitioner, not distributed annotation.
Root-level model evaluation
Logic, correctness, robustness — assessed at the level of reasoning, not surface output. Grounded in deterministic execution facts, scored against frozen, content-hashed rubrics.
Reproducible audit artifacts
Pinned configurations and hashes — a verdict you can re-derive months later, defensible in an audit.
RLHF & preference-data quality
Evaluation of the training signal itself, not only the model's responses.
Dataset QA & annotation audits
Consistency, bias, and coverage gaps across labelled data — found before they train.
Agentic safety & red-team audits
Behaviour under adversarial conditions and open-ended tool use — findings you can act on, reproduced, prioritised.
Every mandate feeds the method.
The method sharpens every mandate.
Operations asks whether a given model is safe and correct; the Research Lab asks how that can be measured and proven at all — two units, deliberately built to reinforce each other.
Can frontier and multi-agent AI systems be evaluated trustworthily, reproducibly, and cheaply on sovereign commodity hardware — a single 32 GB GPU rather than a datacentre?
Data custody you can inspect.
Confidentiality is structural — so the stack itself is the trust signal. Sensitive material is processed on-premises, inside a hardened sandbox, on hardware we own.
Built like the papers:
checkable.
Industry and research under one roof
Operational evaluation access that pure academics lack, paired with methodological depth pure vendors don't offer.
Data sovereignty by design
On-premises processing under strict NDA. The value is demonstrable data custody, not raw compute.
Senior depth, not crowd
Evaluation carried out by a principal practitioner — depth and judgement over distributed volume.
Claims you can check
Every research claim is a falsifiable criterion, labelled measured or predicted — including on this page.
NDA discipline
Confidentiality is structural. Client identities stay private; references remain in the abstract.
Interdisciplinary founder DNA
Deep AI craft paired with a finance- and capital-markets mindset — the "Code → Capital" thesis.
CODE → CAPITAL
From the engine room of model evaluation into a dedicated wealth- and reinvestment architecture.
AI Operations & Research Lab
The evaluation practice and CTC Research Lab's five-workstream program. The senior-evaluation core is delivered today by the founder under mandate for leading AI labs — the company is being formed around that practice.
CTC Advisory
Secure-by-design IT infrastructure, RAG systems, and professional client presences for the DACH mid-market.
CTC Wealth
Reinvestment and wealth architecture — license-compliant in its own regulated structure.
Code to Capital, Inc. is the umbrella entity in formation. CTC AI Operations is the commercial practice — carried today by the founder's senior mandates; CTC Research Lab is the applied-research unit on the same infrastructure.
Two seats, one thesis.
Marian E. Arenskrieger
Owns the evaluation practice, the Research Lab, the transfer of live mandates into the entity, the structural and tax setup, and the long-term capital architecture.
→ arenskrieger.dev · Principal-Investigator profileCommercial & client-facing seat
Active in adjacent AI contracting (voice auditing, AI ad review); builds the go-to-market layer of the Advisory arm and grows into higher-value evaluation work.
Let's evaluate
what's possible.
Open to evaluation mandates and collaborations with leading AI labs. Engagements are scoped under NDA and processed locally.