With the rapid development of artificial intelligence and generative AI technologies, numerous enterprises and organizations are actively leveraging cutting-edge technologies such as natural language processing (NLP) and large language models (LLMs) to create a range of AI-driven products, services, and applications. This article will showcase four enterprises that have achieved significant success in the field of AI innovation and their close collaboration with MongoDB. These enterprises have chosen MongoDB Atlas as their multi-cloud developer data platform, seamlessly integrating operational, analytical, and generative AI data services, thereby simplifying the process of building AI applications.
Based in Australia, Pending AI has successfully developed the Pending AI platform using advanced AI and quantum technologies, aimed at tackling the core challenges of early-stage drug discovery. The platform significantly improves the efficiency and effectiveness of the compound discovery process, enabling researchers to obtain higher-quality and commercially valuable models in a shorter time and at lower costs, thereby advancing the clinical development process.
During the development of core functionalities such as generative molecular design, Pending AI faced significant challenges. This is because the field of chemistry involves an immense number of known pharmacologically relevant molecules, covering over 50 million chemical reactions and billions of molecular building blocks. To precisely design the desired molecules and determine their optimal synthesis pathways, professional scientists often need to undergo a costly and time-consuming trial-and-error process. Therefore, Pending AI urgently needed a database capable of efficiently handling massive amounts of data with outstanding performance to meet the extensive demands of the chemistry field.
图一：待定AI生成式分子设计器等工作成果
After comparing multiple databases, Pending AI ultimately chose MongoDB. As a proven, stable, and easily deployable solution, MongoDB has empowered the Pending AI team to successfully build high-performance deployments on MongoDB Atlas. Particularly as Pending AI began adopting AWS cloud, MongoDB Atlas emerged as a cost-effective fully managed solution, ensuring the lowest latency and security of data transmission by establishing private endpoints between AWS and MongoDB clusters.
Looking ahead, Pending AI plans to further explore the Atlas Search feature in MongoDB 7.0. This move aims to directly integrate currently difficult-to-manage and maintain search functionality into MongoDB, eliminating the dependency on separately maintained Elasticsearch clusters, thus bringing a more convenient and efficient experience to drug discovery.
Eclipse AI: Insight into Customer Interactions for Revenue Growth
Eclipse AI, as a SaaS platform, adds value by transforming customer interaction data scattered across multiple channels (such as customer calls, emails, surveys, product reviews, support tickets, etc.) into profound insights, thereby helping enterprises retain customers and increase revenue. The platform is designed to address the challenges faced by Customer Experience (CX) teams, saving them the time and effort required for integrating and analyzing multi-channel customer feedback data.
In the process of turning customer feedback into actionable insights, Eclipse AI primarily faces the challenge of integrating fragmented customer voice data; secondly, it involves in-depth analysis of this data to extract specific improvement measures to optimize customer experience and prevent customer churn.
MongoDB Atlas, with its flexible document database features, can easily store and index unstructured data embeddings, making it an ideal choice for Eclipse AI. With MongoDB Atlas, Eclipse AI's development team can efficiently and quickly build products while eliminating the tedious work of managing infrastructure. Additionally, features like MongoDB Atlas Device SDKs (formerly Realm) and MongoDB Atlas Search play a crucial role in the functionality implementation of the Eclipse AI platform.
For Eclipse AI, MongoDB isn't just a powerful database; it embodies the concept of data-as-a-service, empowering Eclipse AI to iterate rapidly and release new features continuously to meet market and customer demands.
Since 2015, Safety Champion has been dedicated to revolutionizing the safety management industry, recognizing the paramount importance of workplace safety. Leveraging cloud technology to break free from traditional paper-based processes, the company has been at the forefront of industry transformation. Founder Craig Salter emphasizes the centrality of data in driving the next generation of safety initiatives. Hence, Safety Champion chose MongoDB as its technological cornerstone, adopting MongoDB Atlas in 2017 to enhance cost-effectiveness and reduce administrative burdens.
The user-friendliness of MongoDB facilitates rapid and straightforward application development, significantly boosting performance and saving developers time, allowing them to focus on business innovation and customer needs. MongoDB Charts provide customers with powerful analytical capabilities, enabling evidence-based safety decisions. With nearly a decade of development, especially during the pandemic, the Safety Champion platform has experienced rapid growth, serving over 2000 clients and processing up to 100,000 documents monthly, while doubling its development team size.
图三: Safety Champion Platform
Looking ahead, Safety Champion plans to leverage MongoDB's strengths in generative AI, search, and multi-region capabilities to meet diverse needs. The company is upgrading to MongoDB 6.0, fully integrating MongoDB Search, and intends to utilize MongoDB Vector Search in the second half of 2024. Safety Champion is exploring the use of semantic insights to understand employee text data, combined with large-scale language models to extract valuable information.
Craig Salter notes that customers expect deep analysis, insights, and higher-level meaning from their data. The new Safety Champion platform, supported by MongoDB Atlas, marks the company's move into a new phase, leading the era of safety management with features like generative AI.
Syncly: Accelerating Customer Feedback Analysis Innovation with MongoDB Atlas Vector Search
In today's business environment, rapid response and in-depth analysis of customer feedback have become key drivers of business growth. Voice of the Customer (VoC) services are increasingly complex, requiring AI technology to enhance analysis efficiency. South Korean company Syncly, a startup in the software-as-a-service field, has keenly recognized the potential of the VoC market and launched an AI-driven customer feedback analysis solution.
Syncly's platform integrates multiple channels to collect and manage VoC data in real-time and conducts in-depth analysis through AI to suggest improvement measures for enterprises, enhancing customer relationships. Its core service lies in automatically processing large amounts of data, providing comprehensive visibility into VoC, and emphasizing the role of semantic search in qualitative analysis.
However, traditional search functionality has limitations when dealing with complex data. Syncly actively adopts AI technology to address the challenges of structured and unstructured data, achieving efficient similarity analysis. To this end, Syncly has introduced MongoDB Atlas Vector Search, automating data loading and similarity analysis, reducing developer burden, and increasing productivity.
Figure 4: Syncly Platform
As part of its service expansion, Syncly plans to integrate MongoDB's commercial tools, add search nodes and global clusters to enhance processing capabilities. Additionally, the team will leverage AI to develop VoC (Voice of Customer) services, closely collaborating with MongoDB's Korea team to optimize products and services, ensuring security.
These initiatives will keep Syncly ahead in the field of customer feedback analysis and drive continuous innovation. By harnessing advanced technologies such as MongoDB Atlas Vector Search, Syncly is empowering enterprises to listen to customer voices more efficiently, thus enhancing their business competitiveness.
MongoDB Atlas is a tailored database solution for AI. MongoDB's outstanding capabilities enable enterprises and their development teams to effectively manage rich structured data that does not neatly fit into the strict rows and columns of traditional relational databases, transforming it into meaningful and actionable insights to drive real-world AI applications. Furthermore, the newly added Vector Search feature in MongoDB Atlas allows developers to build intelligent applications driven by semantic search and generative AI, applicable to various types of data. Additionally, MongoDB Atlas introduces the AWS CodeWhisperer coding assistant, providing enterprises with more avenues to explore AI possibilities.
The mentioned functionalities of MongoDB Atlas in the AI domain are just the tip of the iceberg. MongoDB's clientele spans globally, encompassing industries ranging from startups to gaming, automotive, manufacturing, banking, telecommunications, and more. These clients are actively adopting MongoDB Atlas and its features such as Atlas Search and Vector Search, collectively outlining the blueprint for the development of AI and generative AI in the next decade.
