FLOWX.AI

The agent orchestration platform for regulated industries.

FlowX.AI helps banks, insurers, and other regulated enterprises build, run, and govern production AI agents on top of the systems they already have. The platform is LangGraph-native and pairs agent building with a real governance layer — observability, policy enforcement, ROI measurement, and EU AI Act compliance.

This is where we open-source the model side of that work. Everything here targets a real bottleneck in regulated AI — from compliance-grade text understanding to the generation models that power the platform itself.

Available now

CEE PII Detector (cee-pii): multilingual PII detection fine-tuned on GLiNER, weighted for Central & Eastern European languages most detectors ignore. Benchmarked on cee-pii-bench.

Scam & Fraud Detector: scam-guard-qwen17b (1.7B) and scam-guard-qwen06b (0.6B), Qwen3 fine-tunes for scam and fraud patterns. Evaluated on scamguardbench.

Contract Red-Flag Explainer (caveat): on-device model that spots and explains risky clauses in plain language ("what am I actually signing?"). Runs privately in English, Romanian, Polish, and Hungarian, returns structured JSON across a 34-clause taxonomy. Qwen3-4B fine-tune, benchmarked on RedFlag-Bench.

Coming soon

• Ontology Model — generates and maps domain ontologies for agent grounding.
• BPMN Generation (planned) — natural language → BPMN process models.
• UI Generation (planned) — interfaces generated on the fly.

Follow us to get these the day they drop.

flowx.ai · github.com/flowx-ai · LinkedIn