platoseedplatoseed
Open-source multimodal search for AI apps
Morphik is open-source multimodal search for AI apps. We're used by customers ranging from space-tech teams searching across research papers to developers building brokerage agents.
Morphik provides AI teammates for back-office operations, automating AP, billing, collections, and payroll across facilities to reduce manual work. It positions itself as a platform that integrates with existing systems to drive end-to-end automate-and-audit capabilities, specifically for senior living and SNF operators.
Morphik automatically pulls invoices from multiple sources (email, vendor portals, Amazon orders, paper scans, and user uploads), reads and codes each invoice, and posts transactions into the general ledger. It can handle payments and even issue checks. The product connects, reconciles, and automates across the systems customers already run, delivering daily reconciliation, auditable trails, and end-to-end dollar tracking. Deployment begins with workflow mapping and personalization to fit existing systems, approval rules, and constraints, aiming to replace patchwork processes with a single automated workflow.
Who it’s for: Senior living and skilled nursing facility operators with back-office workflows around accounts payable, billing, collections, and payroll.
Customers: IEP Corp
Public case study showing $1M+ annual savings and 150+ facilities live; mentions SOC 2 Type II and ongoing deployment demonstrations
Building Morphik. Prev @MongoDB, worked on scalable database algorithms. Graduated Cornell University in 2022.
AmazonI was one of 3 students in all of India to receive the Tata Scholarship, which allowed me to study Computer Science at Cornell. I dropped out to build Morphik after experiencing terrible RAG and enterprise search first-hand during my co-op at Amazon Robotics. While there, I developed a configuration compiler that sped-up deployment of the Sparrow and Cardinal stations to new locations. I also developed and taught a course in ML Operations and AI as an undergrad at Cornell.
Turn your complex private data into grounded answers
Morphik builds a platform that embeds entire pages, including images and text, and automatically creates a knowledge graph to enable search via UI, SDK, or REST API. It targets knowledge workers who search complex, multimodal documents (diagrams, tables, images) and claims high accuracy for arXiv QA, with configurable extraction rules and enterprise-ready deployment.
Formerly “DataBridge” · why startups rename →
Software tool for designing robots and machines
Touchless AP Automation
