Venture Bytes #112: AI Breaking Through Enterprise Search Barriers

AI Breaking Through Enterprise Search Barriers

In the digital-first modern workplace environment characterized by a proliferation of SaaS tools and knowledge databases, the inefficiency of internal knowledge retrieval is a key challenge. This operational friction results in substantial productivity loss, with the average employee spending over two hours daily searching for internal data and resources. The rise of generative AI-powered enterprise search and knowledge management solutions, like those of Glean Technology, is a direct response to this growing inefficiency.

The scale of the problem becomes even more apparent when looking at industry data. The average enterprise has over 1,000 applications, per Mulesoft (a Salesforce company), of which only 29% actively talk with one another and share data. Compounding this, McKinsey estimates that the average digital worker spends roughly 28% of their workweek managing email and nearly 20% searching for internal information or chasing down colleagues for task-specific assistance. This inefficiency comes at a steep cost, with Fortune 500 companies losing an estimated $12 billion annually due to knowledge work inefficiencies, according to BA-Insight.

The problem of "application sprawl" and data silos is only becoming more pronounced. A Gartner survey highlights that the average desk worker in 2022 was using 11 applications daily—almost double the six used in 2019—with 47% reporting difficulties in locating the information they needed to perform their jobs effectively.

This lack of cohesion between tools and applications creates an environment ripe for disruption, as companies look for solutions that can streamline workflows, eliminate data silos, and improve overall efficiency. Glean Technologies, a leading private company, is poised to capitalize on this opportunity by leveraging AI to optimize operational efficiency and enhance productivity. As enterprises increasingly adopt AI-driven tools to reduce costs and boost output, the appetite for solutions like Glean is only growing.

According to EY, the number of U.S. companies investing $10 million or more in AI is expected to nearly double by 2025. IDC data further supports this trend, predicting that AI spending across key sectors—including software and information services, banking, and retail—will surge to $89.6 billion in 2024 and reach a staggering $222 billion by 2028.

Glean’s platform is already proving its value by helping companies unlock significant efficiency gains and return on investment (ROI). Duolingo, for example, has saved over 500 hours per month through Glean, equating to $1.1 million in annual time savings. Super.com saved over 1,500 hours monthly, realizing an impressive 17x ROI. Webflow reduced its time spent searching for internal data by 300+ hours per month, translating into a 3x ROI.

While Glean is leading the charge, other startups in the enterprise search and knowledge management space are also emerging, further validating the growing demand for solutions that can streamline internal data retrieval and maximize the productivity of knowledge workers. Hebbia, based in New York, offers AI-powered document analysis and search tools for industries like finance and law. Over the past 18 months, Hebbia has grown revenue 15x, quintupled its workforce, and now handles 2% of OpenAI’s daily volume. In early 2024, the company raised $130 million in a Series B round, backed by Andreessen Horowitz, Index Ventures, and Google Ventures, positioning it for rapid growth.

Similarly, Pryon, a North Carolina-based start-up, focuses on AI-powered knowledge management for the tech sector. In September 2023, it raised $100 million in a Series B, reaching a $750 million valuation—a 5x increase from its Series A in 2019. Pryon’s NLP-driven platform enables users to quickly access internal data, significantly improving efficiency.

Quantum Computing Making Tangible Strides

Quantum computing is showing undeniable momentum. While challenges remain—particularly in scaling hardware and reducing errors—the progress in error correction, hardware development, and cloud-based services like Quantum Computing as-a-service (QCaaS) indicates that quantum computing is moving closer to commercialization.

At the heart of quantum computing lies the qubit—a quantum bit that, unlike a traditional bit (0 or 1), can exist in a superposition of states. This allows quantum algorithms to solve complex problems exponentially faster than classical computers. However, today’s quantum systems, known as Noisy Intermediate-Scale Quantum (NISQ) devices, are still plagued by noise and error rates, limiting their usefulness for large-scale real-world applications.

That landscape is changing with quantum computing making tangible strides in both hardware and real-world applications. Since 2018, the number of qubits—an indicator of computational capability—has been doubling every one to two years, a trend likely to continue for the next three to five years.

Earlier this year, Microsoft and Quantinuum, a Colorado-based start-up, claimed to develop the most reliable quantum system on record, achieving an error rate that it claims is 800x better than physical qubits. The two companies ran over 14,000 experiments without errors, pushing quantum computing past its noisy, error-prone stage and closer to practical use. Microsoft also plans to create a commercial quantum machine in collaboration with Atom Computing’s neutral-atom hardware.

While we are moving towards error-free quantum systems, a Google Quantum AI study showed that we might not need fully error-free quantum computers to beat classical supercomputers. They found a stable phase where today's noisy quantum devices can tackle real-world problems in finance, materials, and life sciences. Google also doubled the quantum circuit volume in a benchmark using its Sycamore processor, successfully running complex circuits with 67 qubits and maintaining the same high level of accuracy as earlier tests. This shows that quantum systems can now handle more intricate calculations, even with noise.

Enterprise-grade quantum computing is on the rise, showcasing remarkable advancements. IBM has revolutionized fraud detection with machine learning, reducing false negatives by 5% compared to classical models. QC Ware and Goldman Sachs have outstripped traditional computers in Monte Carlo simulations, a cornerstone for option pricing and risk assessment.

Rigetti Computing is collaborating with NASA to apply quantum machine learning techniques to climate data. D-Wave has partnered with Volkswagen to utilize quantum computing for optimizing traffic flow in urban areas. Google’s Quantum AI team is working with pharmaceutical companies to simulate protein folding processes, which are crucial for drug design.

In a significant collaboration, Crédit Agricole, Pasqal and Multiverse Computing have achieved precise predictions of declining credit scores, equalling the accuracy of its classical random forest model while utilizing 96% fewer initial classifiers. Reflecting the industry's commitment, a BCG survey indicates that 50% of companies invest over $1 million annually in quantum computing, with 70% of these firms sustaining this investment for more than three years.

While big tech is making progress, support from both public and private sectors is also driving the technology forward. Despite a significant 50% decline in overall tech investments, quantum computing managed to secure $1.2 billion in venture capital in 2023. Moreover, public funding for quantum technologies is anticipated to surpass $10 billion in the next few years, indicating strong belief in the sector's potential.

According to McKinsey projections, industries such as automotive, chemicals, financial services, and life sciences are expected to reap the earliest economic benefits from quantum computing. Collectively, these sectors could unlock an astonishing $1.3 trillion in value by 2035. Quantum technology can revolutionize industries by transforming vehicle design in automotive, accelerating material discovery in chemicals, enhancing risk assessment and fraud detection in financial services, and expediting drug discovery in life sciences. Its ability to model complex processes will lead to lighter cars, shorter R&D timelines, optimized portfolios, and personalized treatments.

QCaaS effectively addresses the high costs associated with quantum computing. QCaaS lowers barriers to entry, providing access to users who lack the resources or expertise to develop these technologies independently. It connects users with experts, fosters community support for sharing ideas, and creates opportunities for innovation, including enhancing algorithms and applications. QCaaS offers on-demand access whenever and wherever needed.

Several startups are poised to thrive from the growing adoption of quantum computing. Xanadu, a Canadian QCaaS company, offers both hardware and software solutions, and has formed strategic partnerships with industry giants like Volkswagen, AWS, NVIDIA, Google, IBM, and Microsoft. By focusing on quantum states of light, Xanadu aims to create a modular, networked, room-temperature device with millions of qubits, positioning itself as a potential disruptor in the field.

California-based QC Ware provides quantum computing solutions that are 10 to 100 times faster than those of its competitors. Its credibility is bolstered by partnerships with major firms like Airbus, Goldman Sachs, and JPMorgan Chase & Co., and it has integrated with NVIDIA’s Quantum Cloud to accelerate adoption. The launch of Promethium's Chemical Reaction Suite in September 2024 adds an additional revenue stream, while QC Ware Forge caters to data scientists without quantum backgrounds, broadening its customer base.

Backed by HSBC, Singapore’s National Quantum Office, and JPMorgan Chase Quantinuum holds 108 global patents and over 100 proprietary algorithms. A notable collaboration with JPMorgan Chase demonstrated a 100x improvement over existing industry benchmarks using Quantinuum’s H2-1 quantum computer.

UK-based Oxford Quantum Circuits (OQC) positions itself as a leader in QCaaS, forming strategic partnerships with leading data centers like Equinix and Center Square to enhance adoption. In 2022, OQC integrated the world’s first quantum computer into a colocation data center, and its OQC Toshiko platform is the first enterprise-ready solution, enabling secure global access to quantum computing.