Scribble Data

Scribble Data operates as a machine learning operations (MLOps) product company, established to address the complexities and inefficiencies in preparing data for machine learning and data science applications. The company was co-founded in 2017 by Venkata Pingali and Indrayudh Ghoshal. Their journey began with a shared vision to simplify data for training ML models, stemming from a meeting in 2013 where they connected over the principle that data drives decisions. Pingali, an academic turned entrepreneur with a BTech from IIT Bombay and a PhD from USC, brought deep technical expertise in ML and AI. Ghoshal, with an MBA from the University of Toronto and an Electrical Engineering degree from McGill University, contributed his experience as an operational leader skilled in scaling organizations.

The company's core offering is a platform designed to bridge the gap between data engineering and data science by preparing datasets for continuous use by ML models. This MLOps product provides the foundational blocks for enterprises to build their own models and analyses. Initially, Scribble Data focused on improving trust in data preparation and deployment before narrowing its focus in 2019 to build a feature engineering platform. The company has since evolved, particularly with the advent of generative AI, to develop Hasper, a full-stack applied AI engine that powers AI assistants to transform enterprise workflows, especially in high-value sectors like insurance and pension risk. The business model is B2B, targeting mid-to-large enterprises, with a client base spanning e-commerce, fintech, edtech, and healthcare across North America, Europe, and Asia.

Scribble Data's main product is the Enrich intelligence platform, a modular feature store composed of pre-built feature engineering apps. Enrich is designed to significantly reduce the time-to-market for data science use cases such as creating unified metrics, modeling customer behavior, and developing recommendation systems. It streamlines the data preparation process through versioned pipelines, delivering continuously updated data and providing context through extensive metadata and lineage tracking. More recently, the platform has been enhanced with the Hasper generative AI engine, which combines large language models and machine learning to generate contextual, business-specific insights and predictive recommendations from real-time data. This system allows users to go from raw, unstructured data to an outcome-focused data product quickly, featuring a low-code interface and built-in observability tools to ensure data reliability and security. With offices in Bangalore and Toronto, the company is strategically positioned to serve a global market. A seed funding round in March 2022, led by Blume Ventures, raised $2.2 million to expand the product roadmap and strengthen its presence in North America.

Keywords: MLOps, feature engineering, data preparation, machine learning, artificial intelligence, data science platform, feature store, generative AI, applied AI, data-driven decisions, enterprise data, data reliability, low-code analytics, AI assistants, data product, data lineage, Hasper engine, Venkata Pingali, Indrayudh Ghoshal, insurance AI, pension risk AI, data assetization