4 Essential Steps for Successfully Implementing AI in Your Business

The rapid advancement of artificial intelligence (AI) is expected to significantly impact business operations, with analyst firm Gartner predicting that within two years, half of all business decisions will be fully automated or augmented by AI agents. Business leaders who have already delved into AI share valuable insights at the Snowflake Summit 2025, emphasizing four critical steps organizations should take before fully integrating AI into their operations. 1. Get Your Cloud Strategy Right Wayne Filin-Matthews, Chief Enterprise Architect at AstraZeneca, highlighted the necessity of a robust cloud infrastructure to support AI initiatives. AstraZeneca has developed an AI-enabled research assistant that enhances the productivity of scientific researchers by focusing on the reproducibility of scientific methods and the development of new medicines. The company collaborates with leading academic institutions like Stanford University to run AI experiments. Filin-Matthews stressed that while AI can offer substantial benefits, these can only be realized if the underlying cloud infrastructure is solid. "You cannot be AI-first without being cloud-first," he said. AstraZeneca has also applied AI in commercial areas, automating the creation of marketing materials to serve its complex global market of 126 countries. 2. Focus on Data Governance Amit Patel, Chief Data Officer for Wholesale Banking at Truist, underscored the critical importance of data governance, especially in regulated industries like banking. Patel’s team must ensure data sources are credible, governed, and authorized to comply with external regulations. He discovered that many reliable data sources were lacking internally, necessitating the establishment of a strong data foundation before deploying AI. Additionally, Patel pointed out that employees often underestimate the complexity of implementing AI in an enterprise setting. Setting realistic expectations and defining guardrails around AI usage, such as metadata and quality checks, are essential. "Implementing AI is challenging and requires thoughtful planning and governance to ensure it meets organizational standards," he said. 3. Consider the Quality of Your Outputs Anahita Tafvizi, Chief Data and Analytics Officer at Snowflake, emphasized the importance of maintaining high-quality outputs from AI tools. Snowflake is developing AI-enabled products, including Snowflake Intelligence, which helps business users create data agents. Tafvizi’s team created an AI assistant for the internal sales organization and grappled with questions like whether 95% accuracy is sufficient. Ensuring trust in AI outputs is crucial, and this involves implementing robust governance structures, access controls, and semantic models. "The right balance between innovation speed and governance is key," she said. "We constantly evaluate these aspects to maintain high-quality results." 4. Look for Unanticipated Benefits Thomas Bodenski, Chief Data and Analytics Officer at TS Imagine, a financial technology specialist, discussed the broader benefits of AI beyond mere automation. Since October 2023, TS Imagine has used AI to reduce employee workloads, but Bodenski noted that the technology offers additional advantages. For instance, the company receives 100,000 emails annually from data vendors, each requiring careful review. AI now handles this task, ensuring no crucial information is missed. Previously, this work required the equivalent of 2.5 full-time employees, freeing them to engage in higher-value tasks. AI also fills gaps in service, such as responding to customer inquiries on weekends. "AI has improved our coverage and efficiency while reducing stress and potential errors," Bodenski said. Industry Insights and Company Profiles These recommendations reflect the growing consensus among industry experts that successful AI integration hinges on a solid technological foundation, stringent data governance practices, and a focus on output quality. Companies like AstraZeneca, with its strong emphasis on scientific research, and Truist, a regulated financial institution, are setting benchmarks in this area. Snowflake, known for its cloud-based data warehouse solutions, is pioneering AI tools that enhance user productivity and data management. TS Imagine, a leader in financial technology, illustrates how AI can bring unexpected improvements in operational efficiency. In sum, while AI holds immense potential, its effective deployment requires careful planning and execution. By prioritizing a cloud-first strategy, addressing data governance, ensuring high-quality outputs, and recognizing unanticipated benefits, businesses can maximize the value of AI while minimizing risks.