Platform
Complexity is the Problem
Unleash the Power of Knowledge
Build a massive semantic network.
This vast semantic network made up of real-world entities, including people, organizations, scientific studies, patents, and more, as well as a visual representation of their relationships, already comprises over half-a-trillion semantic tokens and over fifty billion entities (data model objects) and connections between them. And it is growing at an exponential pace.
Reveal unobservable context.
Merging and codifying siloed data and applying data engineering pipelines and Natural Language Processing models, Nostradamus extracts named entities from Big Data, semantically deducing word meaning, and applying high-dimensional embedding and clustering, to discern relevant connections via ontological links. Machine Learning models enrich the graph and each node within it, deducing contextual relationships and metadata not explicitly encoded in the original data sources.
Reveal the Knowledge you Need, Instantly
Identify the ‘unknown unknowns’.
Nostradamus applies semantic search for knowledge discovery, Large Language Models, and high-dimensional vector embeddings, rapidly retrieve knowledge based on understanding the user’s intent, the meaning of concepts and the relationships between them, optimizing the accuracy of the results and making them highly contextual and personalized. This reduces the immeasurable cost of lost opportunities, reliably revealing the ‘unknown unknowns’ the basis of all innovation breakthroughs, and high alpha.
Harness the power of semantic search.
Nostradamus also applies advanced criteria and similarity search techniques. Advanced criteria search is shaped and controlled by a particular business strategy or investment thesis. Similarity search, based on advanced encoding techniques, where the distance between the vector embeddings of companies can be compared and ranked, enables search for similar companies, perhaps competitors or comparables.
Shaping the Future
Solve for Value.
Computing explainable scores via Machine Learning, ALPHA10X probabilistically quantifies previously intangible benchmarks including investment or technology readiness, as well as the likelihood of startup success. ALPHA10X predicts the future via Time Series Analysis and Forecasting, based on Machine Learning and Deep Learning algorithms, and Deep Neural Networks. Sentiment Analysis algorithms process the affective state of data via computational linguistics. We value, simulating the synergy or attach-rate revenue potential of a solution by modelling scenarios in which it is embedded into a larger commercial and technological ecosystem.
Detect Early, Weak Signals.
Nostradamus is designed to pick up weak, early signals, going back to first scientific papers and patent filings on any given topic, identifying emerging trends, and predicting the future with high accuracy. The ultimate goal of Nostradamus is to identify the most valuable technologies and startups three years before they hit the breakthrough point in time in which a range of factors including technology maturity, startup invest-ability, investment volumes, sector leader movements confirm product-market fit for an emerging technology or other type of innovation.