The mortgage industry is well known to be one of the largest but most antiquated industries, ArmorDoc is here to change that. ArmorDoc is laser focused on solving the painful problem of mortgage loan packages that has plagued the industry since its beginnings. At ArmorDoc, data scientists have trained ML models to classify, organize and extract data from a wide variety of forward and reverse mortgage documents, allowing the customer to focus on originating, buying and servicing more loans with the adoption of ArmorDoc’s innovation.
ArmorDoc is the result of the frustration felt by mortgage industry veteran Mike Hartman. Mike previously worked at Lehman Brothers and JP Morgan, pre-financial crisis and at mortgage related financial service and technology companies post-crisis. Throughout his career he’s primarily focused on the quality of mortgage data. The mortgage industry is renowned for being slow to adapt new technologies into their existing workflows with fear of disruption and change.
Mike initially consulted with industry contacts to get a better understanding of the existing document solutions and found it still was an arduous tasks in the industry. The overwhelming response of “yeses”, was enough for Mike to act launching ArmorDoc in 2020. Partnering with a pilot client, a large hedge fund, whom provided ArmorDoc with millions of documents to build and train machine learning models, Mike and the team went into action. After dedicating a full year with the pilot client ensuring the product was executing up to par to meet Mike’s and the pilot client’s standards, the client gave the thumbs up to convert from pilot to roduction. ArmorDoc’s first “deal” involved the processing of over 45mm pages of mortgage documents and knocked it out of the park. So much so, the pilot client is now an equity owner in ArmorDoc.
One thing separating ArmorDoc apart is their willingness to perform “real time demos” with never before seen documents shared by prospective clients just 15 minutes prior to a demo call allowing full transparency in ArmorDoc’s capabilities and allowing prospective clients to compare the initial results to their results after just a few days of retraining models to achieve even more complete and accurate results. ArmorDoc prides itself on being innovative, fast and agile.
Mike initially began detailed research into the existing service providers and their solutions related to indexing documents and data extraction. He found that the current solutions commonly leveraged OCR combined with a rules-driven approach which eventually falls part with unstructured documents, which consequently require human intervention to remedy the problem during post-processing.
In contrast, ArmorDoc uses a completely automated approach. The core difference in the ArmorDoc technique is leveraging the most innovative machine learning and cloud computing technology, which allows incredible flexibility and scalability needed to process unstructured documents in large volumes, but with absolutely no human intervention in the client deliverables. Instead, ArmorDoc’s solution is trained via reinforcement learning, which feeds corrections back into the ML models for retraining until achieving the desired results.
ArmorDoc’s goal is to be the leading provider of document indexing, data extraction, and signature/stamp recognition models for the mortgage industry. According to Mike, ArmorDoc distinguishes itself by solely focusing on three core service offerings. Mike believes the best way to be successful is to be the very best at a few things rather than be okay at a lot of things. Their mission is to be the mortgage industry’s top provider for these three services allowing ArmorDoc to remain laser focused.
Hence, ArmorDoc can satisfactorily serve Lenders, Servicers, and Investors directly, besides integrating themselves into the workflows of other mortgage service providers.
Eventually, this helps clients and partners to streamline their downstream services effectively, leaving the document challenges to ArmorDoc so they can focus on what they do best. ArmorDoc currently provides solutions for forward and reverse mortgages.
ArmorDoc uses a myriad techniques to achieve highly accurate, scalable results which are required for all the challenges mortgage documents present given their wide variety of formats/versions. ArmorDoc’s machine learning models can be effectively applied across a wide variety of loan and document types.
They go above beyond the others implementing the likes of image rotations models to correct upside down or sideways pages, sophisticated logic to identify commingled documents and the ability to expose different versions of the same document.
ArmorDoc is heavily focused on execution and results and does not concern itself with the number of hours needed to get the results. The team at ArmorDoc has complete autonomy thanks to the incredible level of trust and reliability in the company. Empowering the team members helps keep the team motivated and productive and
eliminates the need for micromanagement or setting strict working hours to achieve optimal productivity. The company has adopted a 4.5-day work-week to ensure employees maintain a sound work-life balance unless client deadlines require additional working hours, then it’s an all hands on deck effort to meet, and exceed, client expectations.
ArmorDoc prioritizes customer satisfaction at the forefront of everything they do. It ensures it does its best to obtain as detailed information as possible from customers helping to obtain maximum positive and constructive feedback. ArmorDoc considers transparent, honest, and communicative clients to be the best clients. These clients provide the type of feedback that allows ArmorDoc to grow and constantly improve.
Consequently, this type of straightforwardness allows ArmorDoc to customize their results to meet specified client requirements.
Mike expects 2023 to be ArmorDoc’s hypergrowth year. Adding to its existing client base, creating additional strategic partnerships and integrating with existing service providers platforms creating more streamlined workflows for the entire industry to leverage.