Today, I’m excited to announce the launch of Primer Engines, an integrated suite of industrial-grade natural language processing (NLP) models that bring the power of machine learning (ML) to mission-critical operations at any organization.
Our engines are your flexible building blocks for NLP-powered data processing. Each is pretrained and tuned to work on domain-specific data for use in intelligence and defense operations, financial services, business operations, and more.
You can use Primer Engines individually, or connect them together to build your own custom pipelines that match your organization’s unique workflows. And when you need to retrain them on your own data, you can rapidly do so with Primer Automate — no code or technical skills required.
With Primer Engines, there’s no need to mobilize a team of machine learning experts to build your own NLP solution from scratch. We’re dedicated to continuously pushing the realm of what’s possible with NLP, building new domain and data-specific engines that simply work from day one. Primer Engines are ready to deploy, and each is rigorously trained and tested to be performant, scalable, and fast. Our engines automate complex tasks, such as extracting and linking people, organizations, and locations with the Named Entity Recognition and Linking engines, then classifying them to quickly understand diplomatic relevance with the DIMEFIL engine. The Key Phrase Extraction and Topic Modeling engines identify important information in documents and the Summarization engine automatically generates human-quality reports to share with decision-makers.
Trusted by the U.S. government for military and intelligence applications, Primer Engines unlock the same advanced capabilities for commercial organizations to use in almost any text-related business application. Whether you need to create custom summarizations of hundreds of thousands of customer documents, extract the key earnings phrases from all earnings call transcripts in your portfolio, or identify all topics in a stream of incoming social media posts — Primer has an Engine for you. This means that your teams can automate their reading and writing tasks at scale with human-level precision and discover insights that were previously impossible to find. As a result, anyone in your organization who’s responsible for data analysis, intelligence, or operations can fully access, explore, and take advantage of the firehose of text-based data coming their way every day.
Primer Engines enable you to scale your data analytics operations, leverage richer data insights for better decision-making, and free your teams to focus on their unique expertise. Our engines are:
Primer Engines are powerful, but we’ve also made them exceptionally easy to work with. Simply pick and choose from over two dozen pretrained engines, such as our top performing Named Entity Recognition (NER), to build a pipeline that can power your specific tasks. A plug-and-play API enables fast, seamless integration into your existing applications, tools, or systems, both in the cloud and on premises.
If you need to customize Primer Engines themselves, you can use Primer Automate. With Automate, you can rapidly retrain engines on your own data, or build your own — with no coding or technical skills needed — and effectively teach the engines to be your subject matter expert. Primer Automate reliably takes care of your model infrastructure, from retraining and deployment to monitoring and explainability.
Primer Engines simplify the path to advanced NLP for a wide range of organizations, so they can build and scale their business and operations with confidence. Here are just a few examples:
Financial services: Rapidly uncover trading opportunities hidden in 10-Ks, analyst notes, publicly available research, and news sources.
Business operations: Gain customer and market insights to strengthen brands and improve products and services.
Intelligence and defense operations: Create structure from large volumes of data to reveal insights and accelerate high-stakes decision-making.
Through Primer’s pre-release program, Primer customers have already been building their mission-critical applications on top of Primer Engines. For example:
Cybersecurity provider Anomali uses Primer in conjunction with its own cybersecurity product suite to improve efficiency for internal processes. In one case, it reduced the time it takes to accomplish a task from fifteen to only two hours per week. “We are excited to continue exploring ways to expand our use of Primer technology within our organization” said AJ Nash, Sr. Director Of Cyber Intelligence Strategy at Anomali.
FDB Systems, a financial technology company, tested Primer on billions of finance records against leading cloud and open source NLP solutions. On average, Primer’s out of the box NER model performed 42% better than each of its competitors across all important metrics of precision and recall. Said Simon Mahony, CEO, FDB Systems, “Primer understands our business and is able to train and tune their engines, and other models, as we need in the future.”
Evalueserve, a leading research and analytics solutions provider, has chosen to partner with Primer and use its Summarization Engine to automatically summarize documents in fractions of a second, saving days of work. This enables us to deliver intelligence to our clients faster and makes our platforms more scalable,” said Satyajit Saha, Vice President and Product Lead at Evalueserve."
Our suite of Primer Engines is constantly growing, and we’re releasing new domain- and data-specific engines every month. Going forward, we can’t wait to see the innovative ways that our diverse customers will use Primer Engines to transform their products, operations, and organizations.