Accern: NLP Document Intelligence for Finance
On This Page
Accern built a no-code NLP platform for extracting structured insights from unstructured financial documents, acquired by Wand AI in early 2025 to serve as the data infrastructure layer for agentic enterprise workflows.
Overview
Founded in 2014 and headquartered at 7 World Trade Center in New York, Accern developed a no-code artificial intelligence platform targeting financial teams who need to extract structured signals from unstructured text. The platform's core proposition was accessibility: both technical and non-technical users could build AI-powered extraction workflows without writing code, using adaptive NLP and forecasting models to quantify document content around sentiment, relevance, and entity grouping.
Accern raised $40.3M across multiple rounds, including a $13.3M Series A in December 2020 and a Series B in August 2023, with 56 investors including Allianz Life Ventures and 10X Capital. The company reached 59 employees before its acquisition. In February 2025, Wand AI acquired Accern to strengthen its agentic platform with direct access to billions of structured and unstructured data points. Wand CEO Rotem Alaluf described the rationale: "Wand AI agents will gain enhanced capabilities by extracting and aggregating contextual data, delivering core insights from billions of data events in a comprehensive and actionable manner."
As an independent vendor, Accern competed against financial NLP specialists including Kensho Technologies and Lexalytics, as well as financial document AI platforms like Ocrolus and Daloopa. Post-acquisition, the technology is being folded into Wand AI's agent infrastructure rather than maintained as a standalone product. Buyers evaluating Accern's original capabilities should review the IDP vendor landscape and assess Wand AI's current platform roadmap before committing.
How Accern Processes Documents
Accern's platform ingested unstructured data sources, including news feeds, filings, and financial reports, and converted them into structured datasets through a five-step workflow. The pipeline applied proprietary analytics to quantify extracted content with three primary metrics: sentiment scores, relevance ratings, and entity groupings. This structured output could then feed business intelligence dashboards or downstream models.
The platform used semi-supervised learning, allowing users to expand and refine models over time without restarting from scratch. This iterative approach reduced the time-to-value compared to fully supervised systems that require large labeled datasets before deployment. The no-code interface meant financial analysts could configure extraction workflows directly, without routing requests through data science teams.
Unlike platforms focused on OCR and layout-based document classification, Accern's architecture was text-analytics-first: it assumed digital text input rather than scanned document ingestion, positioning it closer to NLP data infrastructure than to traditional intelligent document processing pipelines.
Use Cases
Financial Services and Asset Management
Asset managers and financial analysts used Accern to monitor news, earnings calls, and regulatory filings for sentiment signals and entity mentions. The platform's relevance scoring helped teams filter high-volume document streams to surface material events. Allianz Life Ventures' participation as an investor signals institutional validation of the use case, though no specific named customer outcomes with processing volumes or accuracy figures have been disclosed publicly.
Compliance and Risk Monitoring
Compliance teams applied Accern's NLP models to monitor unstructured regulatory communications and flag relevant changes. The entity grouping and sentiment quantification capabilities supported risk monitoring workflows where manual review of large document volumes was the primary bottleneck. Post-acquisition, Wand AI positions this capability as enabling "turning information overload into strategic advantage" for finance, asset management, compliance, and risk management teams.
Technical Specifications
| Feature | Specification |
|---|---|
| AI Architecture | Adaptive NLP and forecasting models with semi-supervised learning |
| Processing Approach | Text analytics on unstructured data; not OCR-based |
| Output Metrics | Sentiment scores, relevance ratings, entity groupings |
| Workflow Configuration | No-code, 5-step workflow builder |
| Input Types | Unstructured text: news, filings, financial reports |
| Output Formats | Structured datasets for BI dashboards and downstream models |
| Deployment Options | Cloud |
| Current Status | Acquired by Wand AI (February 2025); technology integrated into Wand platform |
| Total Funding | $40.3M across Series A, Series B, and accelerator rounds |
| Claimed Capability | Converts unstructured text into structured datasets with sentiment, relevance, and grouping metrics (vendor-reported, unverified by independent benchmarks) |
No independent benchmark results or third-party analyst coverage of Accern as a standalone IDP vendor have been identified. All product capability descriptions are sourced from Accern's own documentation and the Wand AI acquisition announcement.
Resources
Company Information
New York, NY, United States. Founded 2014. Acquired by Wand AI in February 2025 and no longer operating as an independent vendor. Prior to acquisition, Accern raised $40.3M from 56 investors including Allianz Life Ventures and 10X Capital. The 59-person team built a no-code NLP platform targeting financial services before the technology was absorbed into Wand AI's agentic data infrastructure.