MassMeds was created out of the need to extract data from thousands of medical-record sets for our own law firm clients. Our firm is handling dockets of more than 2,500 Camp Lejeune claims and hundreds of AFFF-linked disease claims. All of these claims are being worked in house, by our own team. We needed a solution, and MassMeds was created.
Two development shops. One that worked.
MassMeds started more than two years ago, when we hired two separate development shops to create a secure, HIPAA-compliant data-extraction AI tool. For the beta model, one dev team did very well. One did not. We retained the better dev shop and continued building. Their lead AI engineer was MIT-trained in LLM, AI and NLP (and every other acronym for technology that accurately extracts data from records). He ultimately left his dev shop to work full time on MassMeds.
Multiple versions and upgrades followed, including a complex architecture of extraction, pre-processing and micro-prompting across multiple LLMs, plus functionality built for easy human QC and confirmation of found diseases.
More than 10,000 record sets later
More than 10,000 record sets have now been processed through MassMeds, and it has continually produced accurate, relevant disease extractions with source pages fully reported. MassMeds has overwhelmingly beaten human reviewers in third-party testing.
As a plaintiff firm, we completely understand the value of retaining and preserving cash. You pay if and when your client gets paid.