Ricche.ai is a London-based private research and engineering company conducting governed intelligence systems research for financial markets — focused on evidence-conditioned inference, reproducible research, and research-stage decision-support under uncertainty.
Private research laboratory
London, United Kingdom
Financial markets
Active early-stage development
Evidence-conditioned inference
Governance-first, fail-closed
Decisions in these conditions are high-stakes and routinely made on incomplete information. The harder problem is not raw access to data — it is extracting reliable signal from noise, holding outputs to evidence, and reasoning honestly when the underlying state is only partially observable.
Market, event, and contextual information arrives across heterogeneous sources with mismatched schemas, latencies, and reliability.
Real structure coexists with chance correlation, leakage, microstructure artefacts, and adversarial behaviour.
Relevant state is often unobservable directly. Inference must reason about what is hidden, not only what is measured.
Statistical relationships shift. Models calibrated to one regime can degrade silently when the underlying distribution moves.
Without disciplined process, decisions drift toward improvisation under stress. Governance and evidence are the antidote.
Useful patterns exist but rarely surface on their own. Surfacing them requires structured pipelines and methodological care.
Each surface is treated independently — with distinct boundaries, evidence requirements, and supervisory hooks — and integrated under shared governance.
Pattern recognition across price, volume, volatility, order flow, anomalies, and cross-asset relationships. Designed to surface meaningful deviations from noise — and to decline confidence where the evidence does not support it.
Disciplined ingestion, normalisation, cleaning, and structuring of fragmented market data — with attention to provenance, latency boundaries, and reliability classification.
Hypothesis testing, feature extraction, reproducible workflows, and model validation. Built to accelerate structured research without abandoning evidential trails.
Structured analytical outputs with evidence trails, reproducibility, auditability, and governed interpretation. Not reduced to simplistic verdicts.
Compute, ingestion, storage, orchestration, and streaming layers that support the work above without becoming opaque.
The targeted architecture is layered so that evidence, intelligence, supervision, and decision issuance hold distinct roles, boundaries, and accountability. Each layer can be inspected and audited on its own — that separation is what makes the system reproducible and trustworthy.
Market, event, and contextual data arriving across heterogeneous sources with mismatched schemas, latencies, and reliability.
Ingestion, normalisation, and reliability classification that turn raw inputs into traceable, provenance-tagged evidence.
Generates signals, contextual reasoning, and candidate hypotheses — each carrying calibrated confidence and the evidence it rests on. Advisory by design.
Approvals, policy integrity, and readiness certification define the conditions under which the system may act — and hold it closed when certainty degrades.
Human oversight able to inspect evidence, approve, override, and halt. It sits above automation, never beneath it.
The controlled boundary where governed, evidence-bound decisions are issued under supervisory sign-off — traceable and reversible by policy.
Governed intelligence for financial markets is compute-intensive, and the work favours private, controllable infrastructure. Sensitive research and market-data processing are best kept under direct control rather than dispersed across opaque services, so that evidence, models, and audit trails stay coherent and private. The workloads below shape why advanced compute is relevant to the research.
Components below are referenced as part of the development and research environment underpinning the targeted architecture. They are under evaluation or used experimentally, not a claim of production deployment. Some may be retained, replaced, or removed as the research progresses.
Accelerated compute for research and experimentation
Model development and experimentation
Distributed compute under evaluation for research workflows
Workload orchestration in the development environment
Streaming and event tooling under evaluation
Time-series tooling referenced and under evaluation
Cloud foundations for the research and development environment
Approvals, risk posture, override, failure handling, and evidence trails are explicit and reviewable — engineered into the system rather than added afterwards as compliance theatre. The principles below state the posture precisely.
Ricche.ai is in active early-stage development. Public materials are intentionally restrained. Internal dashboards and visuals shown publicly may be preview-oriented or conceptual unless implementation proves otherwise. The website distinguishes between what is conceptual, research-stage, and implemented.
The first stage of Ricche operates as a private internal research environment for market experimentation, monitoring, governed intelligence development, and internal learning. Activity is non-commercial in posture and confined to internal use. There are no external participants and no third-party capital.
Designs and architectures under study. Not yet implemented.
Components under active research, experimentation, and validation.
Components built and operating in a non-production research environment.
Not claimed publicly at this stage. Will be stated plainly when reached.
Classification is conservative by design. Components are labelled by demonstrated maturity, not aspiration.
Active
Under active development
Active
Not yet released
Not available
Not part of current positioning
Ricche is not a recent concept or a quick assembly. It has taken shape through iterative development — foundations first, then architecture, governance, and integration — with each stage refined and revisited before the next is trusted. The public footprint is restrained on purpose; the work behind it is not.
The order in which the work has been undertaken. Stages are revisited as the research matures — the sequence is real; it is not a release schedule.
Framing the problem, the methodology, and the questions worth answering under uncertainty.
Designing the layered separation of evidence, intelligence, supervision, and decisioning.
Making approvals, oversight, evidence trails, and fail-closed behaviour explicit parts of the system.
Bringing the separate surfaces together into a coherent, observable, governed whole.
Continuous testing, validation, and revision — the stage the work permanently lives in.
Publication follows completion, not the calendar — substance before promotion, internal validation before any public claim. A quiet site reflects discipline, not absence of progress.
Not all research, systems, or architecture is disclosed. This protects the integrity of work still under study; it is research hygiene, not secrecy. What is shown is chosen to be honest and durable.
Long-term thinking over short-term noise, continuous refinement over one-off launches, systems over hype, governance over shortcuts, and evidence over narrative.
Ricche runs an active internal research programme across the areas below. Technical notes, governance documents, and reproducibility artifacts are prepared continuously and published selectively. The restraint is deliberate: publication is gated on substance, not on schedule.
Active areas of investigation, each studied with attention to falsifiability, reproducibility, and the operational constraints of financial-market work.
Intelligence that operates under explicit approvals, audit, and supervision — not above them.
Outputs kept tied to traceable evidence, so every claim can be examined and challenged.
Separating durable structure from noise under partial observability and shifting regimes.
Private, controllable compute and inference for sensitive research and market-data work.
Composing operator oversight, governance review, and policy gates without reducing them to theatre.
Hypothesis testing, feature work, and validation that can be re-derived and audited.
Systems that degrade safely when inputs, models, or supervisory state are unhealthy.
Selected technical notes will be published only when they reach a complete and reviewable state.
Evidence that research is real lives in process, not promises. The workflow, principles, output categories, and validation ladder below describe how investigation is run, challenged, governed, and retained — without disclosing any proprietary method, signal, or result.
A single disciplined pipeline carries every line of investigation from raw input to a retained, reviewable record.
A finding is not accepted until traceable evidence supports it.
Outputs must stay understandable and inspectable, not opaque.
Research does not become operational without supervisory review.
Conclusions are re-challenged as new observations arrive.
The kinds of internal studies the programme produces — categories, not results. No strategy, signals, thresholds, or performance are disclosed.
Recurring market structures, examined for whether they persist beyond chance.
How statistical relationships change as market conditions shift.
Whether candidate signals stay consistent under scrutiny and over time.
Operator cognition and decision-support — how findings are framed for human judgement.
Human–AI oversight mechanisms, approval gates, and fail-closed behaviour.
Before a finding is trusted or archived, it must climb the same ladder of escalating scrutiny.
Qualitative status only — no metrics are published while the programme is private.
Active
Multiple
Enabled
In use
Growing
Ricche prioritises truth over narrative, evidence over assertion, and explainability over opacity. Research is governed before it is trusted and challenged continuously as conditions change. The aim is not to look impressive quickly, but to be right durably — and to be able to show why.
Every investigation leaves artefacts. Research is measured not only by its conclusions, but by the records, observations, reviews, and archived evidence it produces along the way. The artefacts below show the durable trail a governed programme accumulates — evidence that work happened and was retained — without disclosing any finding, method, or result.
Initial observations captured for investigation.
Structured documentation of an active investigation.
An assessment of whether evidence supports a hypothesis.
A review of research readiness and oversight requirements.
A retained record preserved for future reference.
Research folded into the growing body of organisational knowledge.
How a single investigation travels from first observation to retained organisational knowledge.
Categories in which work has been undertaken and retained. Categories only — no findings, signals, or results are disclosed.
Research is not discarded once a question is answered. Validated investigations are retained and folded into a growing body of organisational knowledge, so that what is learned in one study remains available to the next. This is how a research programme compounds: through continuity rather than restarts, retention rather than rediscovery, and institutional learning rather than individual memory. The archive is the programme's long-term memory — preserved, reviewable, and built upon over time.
Research value is created not only through discovery, but through documentation, validation, governance, and retention. The objective is durable understanding rather than temporary conclusions — knowledge that survives scrutiny, remains explainable, and is preserved so it can be revisited, challenged, and extended.
Ricche Technical Note 001 — research-stage; advisory only.
Read the full note
Abstract
1. Problem context
2. Definition: evidence-conditioned inference
3. Failure modes
4. Governance requirements
5. Proposed framework
6. Limitations
7. Conclusion
The Evidence Validation Ladder prevents unsupported conclusions by requiring evidence to progress through structured scrutiny before it can influence decision-support or reach research-archive status. It is the operational expression of the evidence-conditioned inference described in Technical Note 001.
Evidence climbs the same ladder of escalating scrutiny each time.
A high-level map of how the governed components fit together. It shows structure and boundaries — not signals, thresholds, model internals, or performance. Every component is research-stage, advisory only, and internal.
Requires evidence to clear escalating scrutiny before it can influence decision-support or be archived.
Observations, evidence source, freshness context, reliability classification.
Accepted-for-research, insufficient-evidence, conflicting-evidence, or archived record.
Fails closed when evidence is weak, stale, or conflicting.
Research-stage · advisory only · internal.
Decides whether a constrained inference output may progress, on the basis of evidence quality and policy.
Constrained inference output and its evidence classification.
Pass, hold, or escalate-for-review.
Decision authority is separated from intelligence generation; no automated escalation.
Research-stage · advisory only · internal.
Bounds expressed confidence to what the evidence supports; declines rather than overstates.
Inference output with evidence quality, freshness, and consistency context.
Evidence-bounded confidence, or an insufficient-evidence outcome.
No unsupported certainty; confidence is never asserted beyond the evidence.
Research-stage · advisory only · internal.
The point at which human oversight inspects the evidence and decides; automation sits beneath it.
Gated output, evidence trail, and governance assessment.
An operator decision, recorded with its rationale.
Operator oversight sits above automation; nothing acts without review.
Research-stage · advisory only · internal.
Retains validated investigations and their evidence trail as durable organisational knowledge.
Reviewed outputs, evidence references, and the governance record.
An archived research record and a knowledge-base contribution.
Auditable trail; records remain reviewable and are built upon over time.
Research-stage · advisory only · internal.
Research-stage · Advisory only · Internal component · No public product access · No investment advice · No performance claim.
How evidence and inference move through the governed system, end to end. This shows operational structure and governance boundaries — input and output types, not the contents of either. Every stage is research-stage, advisory only, and internal.
Classifies incoming data by provenance, freshness, and reliability before any inference runs.
Public data and market context.
Provenance-tagged, reliability-classified evidence.
Untrusted or stale input is quarantined, not silently used.
Research-stage · internal · advisory only.
Binds advisory outputs to the evidence that supports them.
Classified evidence.
Evidence-bound advisory output.
No inference without traceable support.
Research-stage · internal · advisory only.
Decides whether an output may progress, on the basis of evidence quality and policy.
Evidence-bound output and its classification.
Pass, hold, or escalate-for-review.
Authority separated from intelligence; no automated escalation.
Research-stage · internal · advisory only.
Bounds expressed confidence to what the evidence supports; declines rather than overstates.
Gated output with evidence-quality context.
Evidence-bounded confidence, or an insufficient-evidence outcome.
No unsupported certainty.
Research-stage · internal · advisory only.
Human oversight inspects the evidence and decides; automation sits beneath it.
Calibrated output, evidence trail, governance record.
A recorded operator decision.
Operator oversight above automation; no execution authority.
Research-stage · internal · advisory only.
Retains reviewed investigations and their evidence trail as durable organisational knowledge.
Reviewed outputs and evidence references.
An archived, auditable research record.
Auditable; records remain reviewable and are built upon over time.
Research-stage · internal · advisory only.
Research-stage · Internal architecture · Advisory only · No public product access · No execution authority · No performance claim.
This showcase demonstrates system structure and governance boundaries only. It does not disclose trading strategy, model internals, thresholds, signals, or performance.
How research records may be documented, reviewed, and archived within Ricche's governed research framework. The records below are illustrative research-stage examples — chosen to show structure and discipline, not to disclose any finding, method, or proprietary edge.
Why freshness may influence evidence quality.
Review evidence classifications across varying age categories.
Evidence provenance records.
Reviewed under evidence-quality policy.
Further research warranted.
Archived.
How conflicting evidence should be handled.
Evaluate escalation pathways when sources disagree.
Contradictory evidence classifications.
Escalation policy reviewed.
Conflict-handling framework retained.
Archived.
How confidence should be bounded by evidence quality.
Review confidence-expression principles.
Evidence-quality classifications.
Governance gate review completed.
Evidence-conditioned confidence retained.
Archived.
This showcase demonstrates how research records may be documented and archived within Ricche's governed research framework. The records shown are illustrative research examples. No trading strategy, model internals, thresholds, signals, or performance information are disclosed.
Research-stage · Illustrative archive record · Internal framework · Advisory only · No public product access · No execution authority · No performance claim.
A redacted view of the operational state Ricche records inside its governed research environment. It shows the shape of the state and its governance boundaries — never the contents. Identifiers, instruments, signals, thresholds, model internals, execution details, and performance information are intentionally redacted.
Evidence intake
Receives public data and market context.
Active
Provenance-tagged on entry.
Redacted
Evidence classification
Classifies provenance, freshness, reliability.
Active
Quarantine on doubt.
Reviewable
Inference constraint
Binds advisory output to supporting evidence.
Active
No support, no inference.
Redacted
Governance gate
Decides progression on evidence and policy.
Active
No automated escalation.
Reviewable
Confidence calibration
Bounds confidence to the evidence.
Active
No unsupported certainty.
Redacted
Operator review boundary
Human oversight inspects and decides.
Held
Above automation; no execution authority.
Reviewable
Research archive
Retains reviewed records and evidence trail.
Archived
Auditable; reviewable.
Reviewable
REDACTED
REDACTED
REDACTED
REDACTED
Review required
Advisory only
Disabled
Enabled
This showcase demonstrates the type of operational state Ricche records inside its governed research environment. Identifiers, instruments, signals, thresholds, model internals, execution details, and performance information are intentionally redacted.
Research-stage · Internal system state · Advisory only · No public product access · No execution authority · No performance claim.
Answers below describe the company as it currently is — early-stage, private, governance-first. Claims here are kept to what implementation can defend.
Ricche.ai is a London-based private research and engineering company conducting governed intelligence systems research for financial markets. The work emphasises evidence-conditioned inference, reproducible research, and research-stage decision-support under uncertainty.
Governed intelligence is an approach in which an intelligence system's outputs are constrained by explicit supervisory controls — evidence requirements, policy gates, recovery procedures, and operator oversight — so that automation never escapes scrutiny and decisions remain answerable.
Evidence-conditioned inference is reasoning in which every output is tied to the specific evidence supporting it. Confidence is calibrated to what is actually known; outputs decline rather than fabricate when the evidence is thin.
The target domain is financial-market data — price, volume, volatility, order flow, cross-asset relationships — alongside event and contextual streams. Data is used internally for research, experimentation, monitoring, and learning only. It is not redistributed, sublicensed, resold, or provided to external users. Provenance, latency boundaries, and reliability classification are treated as first-class engineering concerns.
No. Ricche is not a public-facing trading platform, retail signal product, or commercial service. The first stage of Ricche operates as a private internal research environment for market experimentation, monitoring, governed intelligence development, and internal learning. There are no external participants, no third-party capital, and no commercial service offering. Market data is consumed for internal research only — not redistributed, sublicensed, or resold.
Not at this stage. Ricche is private and internal by design. Public materials describe posture and methodology; the work is not open to external use.
The development and research stack includes NVIDIA CUDA, PyTorch, Ray, Kubernetes, Redpanda, KDB+, and AWS — referenced as foundations of the targeted architecture, not as a claim of full production deployment.
Governance is treated as part of the system, not decoration. Decision-making authority is separated from intelligence generation; material decisions are evidenced, reviewable, and auditable; operator oversight sits above automation; the architecture is designed to fail closed when certainty degrades.
Five interconnected research surfaces studied as one stack: signal intelligence, market data analytics, research automation, decision-support systems, and data foundations — integrated under shared governance.
Ricche is in active early-stage development. Components are classified honestly as conceptual, research-stage, or implemented. Operational and deployed status is not claimed publicly at this stage; it will be stated plainly when reached.
Ricche operates as a private research and engineering organisation. Public materials focus on systems, methodology, evidential discipline, and governance rather than personalities. Team composition and specialist participation may remain selectively disclosed during active development.
Ricche is founded and led by Manfred Fuss, Founder & Principal Researcher.
Founder of Ricche.ai, focused on governed AI, evidence-based decision systems, and financial-market intelligence research, drawing on more than 25 years of financial-market study.
The work is deliberately systems-led and governance-first; this name is published for accountability, not promotion.
Hypotheses are written down. Questions are framed for falsifiability before answers are pursued.
Implementation is held to what survives scrutiny under load, edge cases, and time.
Supervisory control is structural, not retrofit. Operator oversight sits above automation.
Workflows are designed so results can be re-derived under audit, not only re-cited.
Claims are tied to traceable evidence. Confidence is calibrated to what is actually known.
Clarity about what Ricche.ai is not is as important as clarity about what it is. The following descriptions are not the work being done here.
Infrastructure and strategic enquiries are welcome. The areas below describe the kinds of conversation Ricche is open to — technical, considered, and long-term rather than transactional.
Compute, AI infrastructure, deployment architecture, and scaling.
Research methodology, governance, explainability, and oversight.
Long-term ecosystem and technology direction.
Architecture, governance, and research-review enquiries.
Infrastructure and strategic enquiries are welcome.
The long-term direction of Ricche.ai is intended to come not from noise, hype, or ungoverned automation, but from research that can justify its conclusions, survive scrutiny, and develop capability under disciplined practice.
RICCHE.AI
Governed intelligence systems research for financial markets. London-based private research and engineering.
Materials on this site describe research, methodology, and system architecture under development. Nothing here constitutes investment advice, an offer, solicitation, or public service availability.
Ricche.ai
RICCHE LTD — Registered in England & Wales, Company Number 11483077.
Website revision · 2026.05