An expert consistently building amazing tools that power fact-based decisions: Kensho

The Silicon Review

Kensho is the AI and Innovation Hub for S&P Global. We develop cutting-edge technologies that transform businesses. By pairing the latest advances in machine learning with the unparalleled breadth and depth of data at S&P Global, Kensho gives customers comprehensive, timely, actionable insights for making decisions with conviction. Kensho is headquartered in Harvard Square in Cambridge, Massachusetts and has offices in New York City and Washington, DC. Take a virtual tour of Kensho’s beautiful Cambridge office. Kensho was acquired by S&P Global in 2018, and continues to operate independently, retaining its distinct brand, culture and technology stack.

Kensho leverages S&P’s world-class data to build state-of-the-art machine learning applications both internally for S&P Global and externally for its clients. The company works primarily with natural language data, including complex documents and speech, and build machine learning models that add layers of structure to unstructured and semi-structured data. These foundational AI services improve the breadth and speed of data collection, the depth of data enrichment, and the ease and relevance of data discovery, solving some of the most difficult challenges facing businesses today. Kensho’s solutions transform unstructured data into meaningful and actionable business insights. At Kensho the team insists on action bias and minimizing unhelpful hierarchy and unnecessary process. Kensho collaborates to understand its teammates’ diverse perspectives to solve hard problems together. The company dedicates work time and resources to explore new ideas and to learn new things. Kensho produces technology that is scalable, robust, and useful. The company communicates openly, honestly, efficiently, and with kindness for one another.

Transcribe business audio with unparalleled accuracy

Scribe is specifically optimized for the complexities of your messy, real-world audio and the nuances of spoken language. Trained on more than 100,000 hours of professionally curated audio and corresponding transcripts. Scribe is a valuable tool for any kind of organization or business that deals with high volumes of audio, from the corporate world to academia, from science and medicine to finance and law. Scribe makes your audio accessible and searchable in a snap. While there are many generic, consumer-focused machine transcription services available on the market, Kensho Scribe is unique because it’s purpose-built for business and finance. With a focus on enterprise-scale audio, Kensho Scribe doubles down on accuracy, speed and security using cutting-edge machine learning techniques. Kensho Scribe is optimized to handle the complexities of messy, real-world audio with particular emphasis on numbers, currencies, company names, stock tickers, industry-specific jargon, product names (such as pharmaceutical drugs and tech products), and more. Kensho Scribe can also handle multiple speakers (speaker diarization), heavily accented speech and the nuances of spoken language, including mumbling, stuttering, filler words, hesitation and even self-correction.

Use cases for scribe

S&P Global is a leading provider of financial data, including credit ratings and analytics for global capital and commodity markets. The company catalogs tens of thousands of corporate events and their corresponding metadata every day, including buybacks, guidance, defaults, board meetings and more than 150 other event types. In total, about 40,000 of these events are derived from audio files each year, essentially anything you could dial into or listen to – earnings calls, management presentations, acquisition calls and more.

S&P Global employs hundreds of rigorously trained analysts who have created over 100,000 hours of transcripts of these events over the past decade. The transcripts are 99 percent accurate given Kensho’s strict service level agreements. Kensho’s clients use these transcripts to quickly find the information they need to confidently make decisions. It’s much more efficient to simply search a transcript of an earnings call than it is to listen to the whole event. The organization wanted to streamline the process to free analysts’ time for more value-added activities, but none of the existing commercial speech recognition solutions were up to the task. Too many of the audio clips included poor quality, financial or business jargon, accented speech or other challenges that resulted in errors in the transcriptions. This is how the idea for Kensho Scribe was born.

With a motto of “Try crazy things and make them real,” the Kensho Applied R&D team embarked on the process of building a better machine-based transcription service. The team used the trove of audio and corresponding near-perfect transcripts as a training corpus. More than 100,000 hours of high-quality, domain-specific, labeled data. A dream come true for any machine learning research scientist. One year later, Kensho Scribe was being used to transcribe 99 percent of the earnings calls, management presentations and acquisition call transcripts that S&P Global creates each year. Kensho Scribe saves 1.25 hours per call relative to the legacy process, with more than 50,000 person hours saved to date. The unlocked bandwidth has enabled us to increase transcription coverage by more than 1,500 companies and expand to other high-impact uses like transcribing voicemails, powering compliance efforts, creating meeting minutes and processing interviews. In 2020, Kensho opened the next chapter in the story by beginning to make Kensho Scribe available to clients. As Kensho expands to new use cases for the product, it will continue to push the envelope of transcription quality, evolving the machine learning algorithms that make Kensho Scribe so powerful and accurate.

Meet the leader behind the success of Kensho

Bhavesh Dayalji, CEO has been with Kensho since the early days and held a variety of positions before being named CEO. Prior to Kensho, Bhavesh was a senior leader in Deloitte’s strategy and M&A practice. He was the engagement leader for many high-profile C-suite clients, advising them on M&A strategy, leading due diligence efforts and developing growth strategies in the financial services industry. He started his career at the European Organization for Nuclear Research, CERN, as a software engineer. In his role as CEO, Bhavesh is focused on enabling the team to solve the most complex problems related to structuring unstructured data and helping companies, governments and individuals to make decisions with conviction. Outside of work, Bhavesh enjoys non-fiction books, especially biographies, active sports, Formula One racing, spending time with his family and traveling.