Stride
Stride Technical Briefing
Published June 27, 2023
Supported by IowaEDA and Brale
Overview
Stride.ai co-founder Virajendra Daita walks through the company's unified AI platform for financial institutions — Doc AI for multi-language document processing, Web AI for no-code scraping of regulator and market sources, and an NLU layer for entity extraction and semantic search. The briefing covers an enterprise KYC orchestration that cuts corporate onboarding from ~29 days to ~1 day (saving €2M+/year), an 80,000-docs-per-day remittance processor, and live demos of KYC screening via Refinitiv and structured + unstructured document extraction.
0:00 Introduction to Stride
Virajendra "VJ" Daita introduces Stride.ai — "NextGen AI for Enterprises" — focused on financial institutions and occasional pharma customers, with a unified platform for intelligent automation.
0:26 Platform components: Doc AI, Web AI, NLU
Three pillars of the platform: Doc AI for multi-language and multi-format document processing, Web AI for programmatically tracking regulator registries and markets without code, and NLU for check analysis, entity extraction, and semantic search.
2:13 Workflow editor and pre-built integrations
The no-code build layer with UI and API connectors, intelligent annotation, and pre-built integrations — configuration-first development from requirements gathering through deployed solution.
3:30 Low-resource learning differentiator
Stride's 2019 Dexter paper (dexter.stride.ai) on learning from less data — the entry-barrier-breaker for financial services, alongside on-prem deploy, security and compliance, and tunable customization.
4:48 Enterprise KYC orchestration case study
Corporate KYC reduced from roughly 29 working days to about 1 day using Doc AI, Web AI, and multilingual capabilities. Ubiquitous UX across otherwise fragmented bank tooling, full audit trail, and €2M+ of annual savings.
6:57 Remittance processing at scale
80,000 scanned remittance documents per day processed via pattern recognition. Data-entry and quality-check effort cut by 80%, with seamless integration into existing downstream systems.
7:49 Wedding analyst form standardizer
A simpler case study: a standardized "Wedding Analyst" form captured at >99.9% accuracy, already being integrated into the customer's existing KYC system.
9:06 Live KYC demo
Walkthrough of the KYC orchestration product: pulling screening data from Refinitiv (formerly World-Check), validating against the customer's preferred data sources, then running entity screening in seconds with drill-down into match strength.
11:20 Document processing demo
Live upload of a scanned remittance document. Processed in about 30 seconds with extracted fields rendered alongside the original image, editable to match the back-end schema exactly.
14:02 Unstructured extraction
Final demo showing the linguistic models handling documents with no specific structure, using the core models built over eight years — works for structured, semi-structured, and fully unstructured inputs.
Topics: Identity & KYC, Compliance & Regulation, Developer Tools