
Pulse STUDIO Vision API recently launched!
Founded by Sid Manchkanti & Ritvik Pandey
The founding team has deep machine learning experience at Tesla, NVIDIA, D. E. Shaw, and AWS — as well as research experience at world-class AI labs at Berkeley and Georgia Tech.
Most enterprise data is unstructured, making it difficult to parse with LLMs
Approximately 75% of enterprise data is unstructured, the majority of this is directly within PDF files. This makes it extremely difficult to build RAG applications with this data, and ingestion is often the bottleneck.
Current solutions are slow, inaccurate, and expensive
They personally tested nearly every other tool on the market and found they lack accurate contextual understanding, multi-column PDFs, and multimodal documents. Most of the current technologies are simply wrappers on Textract or Gemini — which have their own inherent flaws.
Pulse STUDIO Vision API, a SOTA document/spreadsheet vision model
The team has trained their own set of Vision Language Models (VLMs) and OCR techniques to bridge this gap. They achieved what they think to be a state-of-the-art (SOTA) vision model for documents and spreadsheets. You’ll get bounding boxes across your documents and spreadsheets, alongside incredible OCR on tables and graphs.
They are also actively working on a novel reasoning tool on spreadsheets using this technology – stay tuned!
