Cactus Life Sciences (CLS) used Microsoft 365 Copilot to rethink how its scientific writers and project managers handle document-heavy, highly regulated work.

Instead of relying on fully manual processes for tasks like search string creation, document review, and structured data extraction, CLS decomposed its workflows and built more than 30 custom automation agents using Microsoft’s agent builder framework.

These agents now help to:

• Retrieve and structure information from scientific literature so writers can focus on deeper analysis and synthesis.
• Support abbreviation checks, formatting consistency, and alignment with regulatory and publication standards.
• Extract and compare insights across multiple documents, while keeping review controls intact.
• Help project managers summarize email threads and generate task lists, improving coordination across remote teams.

CLS emphasizes that AI is there to support, not replace, people. As they put it, “AI supports the labor-intensive parts; it doesn’t own the task.” Human review, quality control, and governance processes remain in place, so Copilot augments expert work rather than automating it end-to-end.

Cactus Life Sciences reports clear, quantifiable efficiency gains from its Copilot deployment and custom agents.

The most notable metric is in structured data work: structured data extraction is now completed approximately 35–50% faster than before, while still going through the same human review and governance processes.

In practical terms, this means:

• Scientific writers can process large volumes of abstracts and articles more effectively, spending more time on interpretation and quality rather than manual data handling.
• Project managers use agents to summarize email communications and generate task lists, which helps them manage overflowing inboxes and coordinate distributed teams more efficiently.

Beyond the numbers, CLS is seeing a growing culture of innovation. A Copilot Champions community and peer-led experimentation are helping teams continuously find new ways to apply AI and automation to everyday work.

Security and governance were non-negotiable for Cactus Life Sciences when adopting Microsoft 365 Copilot, given the sensitive and proprietary nature of pharmaceutical and clinical data.

CLS chose Microsoft 365 and Copilot because of their enterprise-grade security features, including:

• Built-in protections that help keep client data secure.
• Enterprise-grade authentication to control access.
• Dedicated project environments so teams can work confidently with sensitive information.

On top of the platform capabilities, CLS maintains a Human Anchored AI approach:

• All AI-enabled outputs remain subject to established human review, quality control, and governance.
• Quality and governance processes have not been relaxed to accommodate AI; they are unchanged to address concerns around hallucinations, bias, and error tolerance.
• A Copilot Champions program and a dedicated training team ensure employees understand both the capabilities and limitations of AI tools.
• A centralized repository for prompts, agent ideas, and best practices supports consistent, responsible use.

CLS’s stance is clear: technology is there to support people, not replace them. As they continue to evolve their use of Copilot and agents, the focus remains on secure, well-governed AI that helps teams deliver high-quality medical communications for clients.