Theia Insights secures $8 million to develop an alternative to traditional static industry classification systems.

      The AI firm based in Cambridge, started by a former Amazon Alexa research scientist, has developed a self-learning economic map that represents companies as multidimensional entities rather than categorizing them into a single classification. MiddleGame Ventures led the funding round, while Unusual Ventures participated again.

      Amazon is not just a retailer; it also functions as a cloud computing service, logistics operator, advertising network, media entity, and hardware producer. The traditional classification of Amazon under Consumer Discretionary fails to capture its full scope.

      The Global Industry Classification Standard (GICS) and Industry Classification Benchmark (ICB) systems used for building portfolios, constructing indices, and measuring risk were created in a different time concerning corporate structures and have not significantly evolved since then.

      Theia Insights, a deeptech company in Cambridge, has dedicated four years to creating an alternative system and has raised $8 million in Series A funding to further its development.

      This funding round was led by MiddleGame Ventures, with contributions from Further Ventures and Unusual Ventures, which also spearheaded Theia's previous funding round. The total amount raised to date is now $14.5 million.

      Founded in 2022 by Dr. Ye Tian, a former PhD research scientist at Amazon Alexa with expertise in NLP and AI, Theia has assembled a team with experience from Nasdaq, Morgan Stanley, Meta, UC Berkeley's economics department, and the computer science department at the University of Cambridge.

      Theia’s primary technology processes regulatory filings, earnings calls, press releases, and financial data from publicly traded companies, using proprietary NLP and quantitative models to extract business activities and create what it refers to as a dynamic, self-learning ontology of the global economy.

      In contrast to static classification systems, it views companies as multidimensional entities, monitoring "who does what by how much" across all revenue streams as they evolve over time.

      This mapping supports four products: a Dynamic Industry Classification system (TIIC), a Concept2Universe tool (C2U) that translates investment themes into data-supported company universes, a Thematic Factor Model (TFM) that identifies the structural trend drivers behind stock price fluctuations, and Theme Watch Indices (TWI) for real-time tracking of global industry trends.

      Theia's current clientele consists of institutional players, including prominent index providers, major banks, large asset managers, and hedge funds. The strategic purpose of the new funding is to expand into private markets, an area lacking a comparable dynamic classification system, which is becoming increasingly important as more institutional capital is allocated to unlisted assets.

      This funding will also enhance engineering and research capabilities and support commercial scaling.

      Patrick Pinschmidt, co-managing partner at MiddleGame Ventures, characterized the issue as both structural and timely: “Financial markets continue to depend on static classification systems that have evolved little over recent decades. Theia’s methodology develops a dynamic, AI-driven map of a company, sector, or investment theme, offering transformative tools for investors and AI systems to analyze.”

      The focus on AI is intentional: as financial institutions create their own AI workflows for research, portfolio construction, and capital allocation, they require structured economic data that machines can interpret, not solely what analysts can understand. Theia aims to position its ontology as essential infrastructure for both.