The problem
Around 5,000 solo advisors, half their week lost to compliance paperwork.
France counts roughly 5,000 AMF-registered CIFs, most of them solo practices. They operate under MiFID II with the compliance surface of an institution: suitability assessments, adequacy reports, regulatory justification narratives, a 10-year audit trail. More than half of their working time goes to that paperwork instead of advising.
The tooling available to them is CRM-shaped: client files, task lists, document storage. The regulated work itself, the analysis and the writing, stays manual.
Socle started with deliberately non-leading interviews with practicing CIFs, and it corrected my assumptions more than it confirmed them. While I expected reluctance about AI, on regulatory grounds if nothing else, they were already running AI agents with financial-lawyer and accountant personas, anonymizing client data by hand, and scraping fund valuations from Yahoo Finance with AI-generated Python scripts, because the professional data feeds only pay off above roughly €10M under management. Their constraints were elsewhere. Client data could never reach an AI: “tout ce qui est public, je bombarde,” as one of them put it, everything public goes in, client files never. And when asked what would actually help, the answer was not a dashboard: “si j’ai pas une assistante, il faut une IA.”
Those constraints are the product. CIFs don’t need another dashboard. They need a tool that understands their regulatory obligations, handles the paperwork so they can advise, and leaves strategy where they insisted it belongs: with the human. That last part goes deeper than preference. The CIF is personally accountable for the advice they give. An AI cannot be accountable.
What was built
A back office where the AI does the regulated paperwork and the advisor keeps the authority.
Socle consolidates fund universe management, client onboarding, portfolio construction, adequacy analysis, and AMF-compliant document generation into a single encrypted workspace. One design principle runs through every feature: AI proposes, screen controls, CIF confirms. The AI pre-fills suitability assessments, suggests portfolio allocations, drafts justification narratives. The CIF always reviews, edits inline, and explicitly confirms before anything takes effect. This is also the human-oversight posture the EU AI Act expects of AI in consequential decisions, built into the workflow rather than added as a checkbox.
The assistant comes as personas, and they work in both directions. Default personas cover the standing needs (financial analyst, compliance expert, legal advisor) and the CIF can define their own. They give second opinions on strategy without owning it, and they watch the CIF’s own work too: a persona raises an alert when a draft drifts toward non-compliance, away from the client’s risk profile, or past a threshold the CIF has set. For a solo practice, that is the second pair of eyes the advisor cannot hire.
Design decisions
The four calls that shaped the architecture.
No client data reaches an external API without anonymization. Every user message passes through a two-layer detection pipeline before reaching the LLM. Layer one: multilingual NER (distilbert-multilingual-cased-ner-hrl, covering 10 EU languages) catches names, organizations, and locations. Layer two: regex patterns catch IBAN, NIR (French social security), email, phone, postal codes, and dates of birth. Detected entities are replaced with consistent global tokens ([PERSON_1], [PORTFOLIO_1]) via a registry that persists across the session, and responses are de-anonymized client-side after return. The audit trail records both the anonymized prompt and the restored output.
GDPR erasure and AMF retention pull in opposite directions, so the data model splits. Client data separates into non-PII (MiFID profile, risk tolerance, investment objectives) and client PII (names, contact, SSN), inside one SQLCipher-encrypted database. Right-to-erasure deletes the PII while the anonymized profile survives for AMF’s 10-year retention requirement. Both regulators are satisfied by the same schema.
Suitability is decomposed into structured scoring, grounded in actual fund data. Adequacy assessment covers seven dimensions (investor type, knowledge level, risk tolerance, loss capacity, investment horizon, objectives, sustainability preferences), each scored pass/fail/partial against the fund’s EMT target-market data. The AI’s compliance narratives are grounded in that scoring and in the underlying fund data, and every generated document records its generation context: timestamp, data snapshot, model version, who reviewed and approved. AI-generated content stays visually distinct from CIF-edited content, down to the badge.
The deliberation is part of the record, not just the outcome. Every AI conversation is a persistent thread that can fork. Branches are preserved and dated, and the trail feeds the final recommendation: what was explored, which personas weighed in, what was set aside. This is not best-execution evidence; it is suitability evidence, proof that the recommendation followed from the client’s profile and needs, with the alternatives considered on the record, which is what the AMF expects an advisor to be able to show. The “Fork as branch” control in the proposal screen above is this decision surfacing in the interface.
What this validates
That AI-assisted compliance can respect the advisor’s authority and still hold up to inspection.
- Multilingual PII anonymization can be fast enough for interactive use (NER + regex in the request path)
- MiFID II suitability can be decomposed into structured, per-dimension scoring rather than narrative-only assessment
- AI-generated compliance documents can carry audit metadata that satisfies AMF inspection requirements
- A single encrypted SQLite database can serve as both the working data store and the GDPR-compliant archive
- Default personas (financial analyst, compliance expert, legal advisor) can act as guardrails, alerting when a draft drifts from compliance or the client’s profile
- Conversation branching can double as the suitability record: alternatives considered, preserved, and traceable into the final recommendation
Limits
What stands between this prototype and production.
Socle needs to ship as a standalone local application, so that installation and execution are one step rather than a developer setup. Local-first packaging creates its own obligation though: the CIF becomes accountable for backing up their data and rolling it back when something goes wrong, and the current system does neither. Backup and rollback are the last step before this could responsibly be put in an advisor’s hands.
The second gap is the inference route. By default Socle runs through the Claude Code SDK rather than the metered API, which keeps prototype costs near zero. There are fallback routes to the Claude API and a standard interface for plugging in any LLM, but the default path inherits every behavior change in the SDK, message streaming being a recent example.
Socle is one half of a pair. The RWA allocation gateway prototypes how tokenized assets reach retail portfolios; Socle is the regulatory back office of the advisor doing the allocating. Same workflow, seen from both ends.