Heping Liu, the Senior Machine Learning Principal at Workday, whose journey from academia to the pinnacle of technological innovation is a testament to resilience, strategic foresight, and an undying passion for artificial intelligence. In the world of enterprise software, where innovation drives success and adaptability defines longevity, Workday stands out as a beacon of excellence. With a mission to empower organizations to adapt to a changing world, Workday leverages innovative technology to deliver cloud applications for human resources, finance, planning, spend management, and analytics.
Journey to Workday
Heping Liu’s journey as a technology leader is rooted in a rich tapestry of academic and entrepreneurial experiences. Beginning with a master’s degree thesis, Dr. Liu explored the potential of neural networks to forecast financial indices and used the Markowitz model for portfolio optimization. This early exposure to the power of computational intelligence laid the groundwork for his pursuit of a Ph.D. in 2004, where his focus was on advancing the field of AI. Transitioning from academia to industry, Dr. Liu embarked on a challenging yet rewarding path through several startup companies, driven by the dream of establishing his own venture and eventually taking it public.
The entrepreneurial landscape offered Dr. Liu invaluable insights into the nuances of building and leading a company. In Beijing, he founded Unigroup AI Technologies Inc., where he not only managed the company but also spearheaded the development of advanced AI technologies and products. This experience underscored the critical importance of assembling a cohesive team, navigating the complexities inherent to startups, securing funding, and remaining at the forefront of technological trends. These lessons became integral to Dr. Liu’s leadership style, characterized by strategic planning and a keen ability to surmount challenges.
Joining Workday Inc. marked a new chapter in Dr. Liu’s career, presenting the formidable task of integrating cutting-edge AI solutions within an established enterprise framework. He successfully proposed, initiated, and led several pivotal projects that were showcased at the Annual Workday Product & Technology Conferences, known as the Annual Spelunking Conference. These initiatives, such as the Workday Resource Forecasting/Optimization Platform, Forecasting-as-a-Service and a conversational front end for the Workday system using large language models (LLMs), exemplify Dr. Liu’s visionary leadership and his capacity to overcome technical and strategic hurdles.
Workday’s Mission and Vision
Workday, Inc. is a leading provider of enterprise cloud applications for finance and human resources. Its mission is to deliver innovative cloud-based solutions that empower businesses to thrive in the digital age. By providing user-friendly applications that streamline operations, Workday enables organizations to reach their full potential. Leveraging AI and machine learning, the company seeks to revolutionize how businesses manage human resources, finance, and other critical functions.
Workday envisions a future where every organization has access to the tools and resources needed to succeed in a digital-first world. The company aspires to be at the forefront of digital transformation in the enterprise software industry, offering advanced technology that enhances efficiency, collaboration, and growth. Workday’s vision is to empower businesses to make data-driven decisions and seamlessly integrate technology into all aspects of their operations, fostering a more connected and productive workforce.
Building and Managing Effective Teams
A key component of Dr. Liu’s leadership philosophy is building and managing motivated and effective teams. He places significant emphasis on fostering an environment where team members are self-driven, disciplined, and aligned with both the technological and cultural aspects of the organization. Clear communication is paramount, ensuring that each team member understands their role and the broader organizational objectives. Dr. Liu believes in granting autonomy to his team, allowing them to explore creative solutions while offering guidance and support as needed. This approach cultivates a culture of collaboration, respect, and continuous learning, where team members are encouraged to take ownership of their projects and grow their skill sets.
Hiring the right talent is critical to Dr. Liu’s management strategy. Beyond technical skills, he prioritizes cultural fit. Mentoring and professional development are integral to Dr. Liu’s approach, enabling team members to thrive and innovate. Recognizing and celebrating achievements further enhances morale and motivation, creating an environment where individuals feel valued and driven to excel.
Long-Term Vision and Goals
Dr. Liu’s long-term vision involves continuing to drive innovation in areas such as forecasting, optimization, machine learning, generative and multimodal AI, and Artificial General Intelligence (AGI). His focus is on creating impactful solutions that have the potential to transform industries. A key aspect of this vision is advocating for and building a data environment that is friendly to generative and conversational AI. Utilizing distributed system design, Dr. Liu aims to develop cutting-edge products and services based on generative and multimodal AI models, while exploring the application of time series foundation models to various types of time series data and optimization models.
Within Workday, Dr. Liu’s future goals are centered around intelligent solutions by advancing time series forecasting and optimization models, enhancing distributed systems, and exploring and developing generative and conversational AI applications. By integrating advanced AI and machine learning techniques, he seeks to make Workday’s HR and financial data enterprise solutions more intelligent and adaptive. Leveraging generative and conversational AI, Dr. Liu aims to create innovative tools that enhance user generative and conversational experience and provide deeper data insights. He is committed to advocating for a domain-friendly data environment that supports efficient use of generative and conversational AI, ensuring scalability, reliability, security, and performance of AI-driven products and services in the big data environment.
Workday Culture and Values
Workday’s culture is centered around the belief that happy employees lead to happy customers. The company fosters a collaborative, inclusive, and innovative work environment. Workday prioritizes employee well-being, diversity, and inclusion, and emphasizes a culture of continuous learning, open communication, and mutual respect.
Workday’s values encompass several key aspects: employees, customer service, innovation, integrity, fun, and profitability. The company is committed to fostering a supportive environment where employees are encouraged to excel and grow, with a strong emphasis on respect and continuous improvement. Dedicated to delivering exceptional value, Workday prioritizes customer satisfaction and success. The company embraces innovation, learn from outcomes to remain competitive, and deliver long-term value. Honesty and accountability are central. Workday also promotes a fun and enjoyable work atmosphere. While profitability is important, the company prioritizes living its values and making sustainable financial decisions.
The Future of AI
As the AI landscape continues to evolve, generative AI and large language models are poised to revolutionize various industries by transforming how data is interacted with, processes are automated, and insights are generated. The potential emergence of Artificial General Intelligence (AGI) represents a significant milestone, with AGI systems anticipated to possess broad cognitive abilities across diverse tasks and domains. According to OpenAI, AGI refers to highly autonomous systems that can outperform humans in practically all economically valuable work, encompassing levels of interaction, problem-solving, and innovation that surpass current AI capabilities.
At Workday, AI is deeply integrated into the company’s platform, driving intelligent predictions and automation. The company is actively preparing for the opportunities presented by advancements in AI, committed to incorporating the latest technologies to enhance its services and maintain a leading edge in innovation. By embracing the potential of generative AI, Workday aims to continue delivering cutting-edge solutions that empower its users and transform industries.
Transforming the Tech Industry
Given the opportunity, Dr. Liu would focus on the early development of generative AI and the integration of function calling with large language models (LLMs). The potential of generative AI spans various applications, enhancing AI’s ability to generate high-quality content, design innovative solutions, and assist in creative processes across multiple domains. Integrating function calling with LLMs significantly enhances their usability and effectiveness, allowing LLMs to execute specific tasks, access external databases, and interact with various software systems seamlessly. This integration leads to more intelligent and context-aware AI agents, improving user experience and productivity.
Impact on AI
Dr. Liu has significantly impacted the field of artificial intelligence through his contributions to advancing AI technologies, developing products across various domains, and educating young professionals about AI. His research has resulted in the publication of approximately 20 papers in academic journals, presenting innovative AI, computational intelligence, and deep learning algorithms and models with applications in investment and pricing, revenue management, energy, healthcare, and nano-manufacturing.
In the industry, Dr. Liu has led the development of AI-driven platforms, products, and services across multiple companies. His initiatives include creating time series price forecasting and supply chain optimization models for the food industry, developing a next-generation analytics engine for online advertising, and pioneering big data analytics and clustering modeling for IP intelligence and reputation. These contributions have been utilized in advertisement targeting, transaction fraud prevention, and marketing industries, showcasing Dr. Liu’s ability to drive innovation and deliver tangible business value.
In 2018, Dr. Liu founded an AI technology company, leading teams to build a conversational product for the finance and investment industry using Natural Language Understanding (NLU) technologies. The product’s major components include a conversational platform based on NLU technologies, a knowledge graph platform utilizing Neo4J and APOC and graph algorithms, a search platform based on ElasticSearch, a big data ETL platform using Spark, Kafka, and Redis, and a web crawling platform to collect real-time finance text and quantitative data.
At Workday, Dr. Liu has proposed, initiated, and led several key projects, presenting them at the Annual Workday Product & Technology Conferences, known as the Annual Spelunking Conference. These projects and proposals include the Workday Resource Forecasting/Optimization Platform, Forecasting-as-a-Service, a conversational front end for the Workday system using large language models (LLMs), and transforming the Object Management Service (OMS) into an AI-Agent friendly data environment. Dr. Liu’s contributions to AI continue to drive innovation and enhance Workday’s ability to deliver intelligent, adaptive enterprise solutions.
In addition to his professional work, Dr. Heping Liu dedicates time to mentoring and teaching young professionals, helping them learn the latest advancements in AI. His commitment to education and mentorship underscores his dedication to fostering the next generation of AI leaders, equipping them with the skills and knowledge necessary to navigate and contribute to the dynamic field of artificial intelligence.
The Future of Forecasting and Optimization
Over the next five years, machine learning is poised to play an increasingly central role in forecasting and optimization platforms. Advances in generative models will enable more generalized predictions and optimized decision-making processes, driving efficiency and innovation across industries. A significant trend in forecasting has emerged, with companies leveraging transformer models to build generalized large time series foundation models. Examples include Amazon’s Chronos, Salesforce’s Moirai, Google’s TimesFM, and Datadog’s Toto.
These large time series pre trained foundation models represent a paradigm shift in forecasting. Traditionally, developing an effective forecasting model required a forecasting expert to tailor a model to a specific dataset. However, with these generalized models, this expertise is no longer a necessity. The models can generalize across various datasets, democratizing prediction modeling by reducing the reliance on specialized forecasting expertise and the need for specific data and computational resources.
AI/ML and Big Data Analytics: The Next Big Trend
The next big trend in AI/ML and big data analytics is expected to be in generative and multimodal AI and AGI. These technologies are set to continue evolving, with some models becoming much larger and others, oriented to specific industries or applications, becoming smaller. Equipped with either large or small language models, the application of AI agents is anticipated to become widespread.
Industry leaders, such as ChatGPT 3.5 and subsequent versions, serve as language foundational models, and companies are exploring ways to apply these foundation models to their businesses. AI agents based on these foundation models have demonstrated the potential to connect LLMs with company domain knowledge, providing automated services. AI agents act as cognition amplifiers, anticipating user needs and helping them accomplish tasks. They offer predictive, conversational and generative capabilities along with advanced analytics, providing users with intelligent and context-aware interactions.
The internet may evolve into a network of AI agents, with humans focusing more on reviewing and approving the work of AI agents. With the introduction of LLMs, user interaction with data is shifting from customized user web interfaces to conversational approaches. This revolutionizes how users interact with proprietary data, enabling dialogues, report generation, dataset comparison and analysis, model building, and actionable insights. The interaction between users and their data becomes more intelligent, leading to a trend where user needs shift towards business intelligence as users have more freedom to ask complicated analytical questions.
Both AI agents and conversational interaction require a friendly data environment and smart function calling, which are challenging to build. Continuous improvement of model accuracy is necessary, but understanding domain-specific data remains a significant challenge. To assist AI agents and conversational interaction, labeling and tagging data will become critical to create a more friendly data environment.
Workday recognizes these industry trends and opportunities. The company has partnered with Salesforce to develop a new AI employee service agent that will automate time-consuming tasks, provide personalized support, and surface data-driven insights to help employees work smarter and faster. The combination of Salesforce’s new Agentforce Platform and Einstein AI with the Workday platform and Workday AI will enable organizations to create and manage agents for various employee service use cases. This AI agent will work with and elevate humans to drive employee and customer success across the business. At Workday, AI is at the core of the platform, powering intelligent predictions and automation like no one else can.
Advice to Aspiring Entrepreneurs
To budding entrepreneurs aspiring to venture into the AI sector, Dr. Heping Liu offers valuable advice: stay curious, be resilient, and prioritize ethical considerations in your work. Understanding your target market, staying attuned to industry trends, and being prepared to pivot when necessary are critical components of success in the dynamic and rapidly evolving field of AI. Continuous learning and adaptation are essential, as the AI landscape is constantly changing and presenting new opportunities and challenges.
Furthermore, Dr. Heping Liu emphasizes the importance of ethical considerations in AI development. Strive to develop AI solutions that are fair, transparent, and beneficial to society. By prioritizing ethical considerations, entrepreneurs can ensure that their innovations contribute positively to the world and build trust with users and stakeholders. This approach not only enhances the impact of AI solutions but also aligns with the broader societal goal of harnessing technology for the greater good.
