Artificial intelligence · Banking & specialized finance
Industrial AI. Sovereign. Measurable.
We design and deploy agentic AI platforms for organizations that demand traceability, compliance, and performance — from regulated banking to advanced industry. No dead-end POCs: production systems, operated on European infrastructure, aligned with ACPR, DORA, and EU AI Act requirements.
What we have built
Each addresses a specific operational pain point at a regulated player. No demonstrators: multi-agent architectures integrated into existing IT systems, measured in time saved and risk reduced.
01 — Commercial pre-scoring
A platform that continuously scans the economic landscape, identifies companies eligible for financing, and composes ready-to-send commercial proposals — with tax filing analysis, multi-criteria scoring, and personalized emails.

02 — Recovery
A platform dedicated to complex receivables portfolios: lots, contracts, multiple lessors, actuarial and linear schedules. The AI detects anomalies, proposes recovery actions, and industrializes large-scale follow-up.

03 — KYC / AML
A multi-agent architecture covering the full KYC cycle: onboarding, periodic review, risk detection, and regulatory reporting. Designed for two frameworks — ACPR / TRACFIN in Europe, BAM / UTRF / GAFIMOAN in the Maghreb.

04 — Credit & Leasing
An agent chain that transforms a raw multi-company dossier into a committee-ready report. OCR, classification, financial extraction, reputation research, aggregation, and structured writing: all in a single execution.

Our approach
We don't sell a model — we deliver a complete chain: ingestion, reasoning, control, and output. Every brick is observable, logged, and operated by us.
One agent per business task, orchestrated by a supervisor. This specialization produces traceable, auditable, and easily improvable results — where a monolithic model remains a black box.
Models reason on your own corpora: internal procedures, regulations, contracts, decision history. Responses are sourced, updatable without retraining, and stay within your perimeter.
Everything is deployable on-premise or in a European private cloud, with an air-gap option. Sensitive data never leaves your environment. DORA and EU AI Act compliance built in from the start.
Beyond banking
Our agent chains, traceability standards, and sovereign stack apply wherever AI must be industrial, measurable, and auditable. Here are the domains where we are currently engaging client use cases.
Compliance by design
Working with credit institutions requires a framework. Our platforms are designed to fit the standards that structure your business, and we document every flow for your internal and external auditors.
Our fields of intervention
We work with credit institutions, lessors, and specialized finance players — on projects where AI must navigate strong regulatory constraints and existing structured IT systems.
Let's talk
We start with a short 5-to-7-week scoping phase to qualify value, target architecture, and the path to production. No empty promises: a decision-making deliverable at the end.