AI-powered document fraud detection for lending and KYC workflows
Resistant AI detects forged and manipulated documents across banking, lending, and identity verification use cases. The stack—Python, FastAPI, AWS Lambda, DynamoDB—reflects a microservices-oriented backend tuned for high-throughput API inference. Hiring is engineering-heavy with five roles open, and the pain-point list reveals active work on model drift monitoring and synthetic-identity detection, suggesting they're moving beyond static fraud patterns toward continuous model evaluation and account-takeover prevention.
Resistant AI builds a document fraud detection platform deployed in digital lending, KYC onboarding, merchant screening, and employment verification workflows. The system analyzes documents across all formats and languages—PDFs, images, bank statements, IDs, pay stubs—and renders a fraud verdict in under 20 seconds via manual upload or automated API integration. The company is headquartered in Prague and operates with a 51–200-person team split across engineering, data, marketing, and research functions, with active hiring in Czechia and the United States.
Core stack: Python, FastAPI, AWS (Lambda, DynamoDB, Aurora, API Gateway, Step Functions, Batch), Docker, GitLab, Vue.js. Actively adopting AWS CDK for infrastructure-as-code.
Bank statements, pay stubs, IDs, utility bills, tax forms, and other document types in all formats (PDFs, images) and languages, with a target accuracy of 99% on fraud verdicts.
Other companies in the same industry, closest in size