Posted in

From prompts to partnership: How LTM’s Rajesh Kumar colla…

by ailona•

LTM CIO Rajesh Kumar evolved Microsoft 365 Copilot use from simple email prompts to complex Researcher queries and low-code agents, driving companywide AI adoption, improving productivity, enabling strategic research, and building agents for talent matching and workflow automation.

Rajesh Kumar’s use of Microsoft 365 Copilot has shifted from simple prompts to structured, multi-paragraph instructions. His team at LTM now integrates Copilot across workflows and strategic research. This change moves Copilot from an assistant into a collaborative agent for enterprise tasks.

Main feature/change and impact

Kumar shifted prompt design from one-liners to detailed, multi-step instructions stored in OneNote. Copilot’s Researcher agent now performs deep vendor analysis, product comparisons, and risk assessments. Embedding Copilot into Teams and Outlook preserves context and removes copy-paste overhead. The net impact is reduced interruptions, faster decision cycles, and more reliable starting points for strategic reports.

Practical implications

LTM adopted department-specific Copilot sessions and hackathons to scale adoption. Non-developers use Copilot Studio to build low-code agents tailored to business needs. Agents automate talent matching using ERP data and digital resumes, speeding staffing decisions. Operationally, users get consolidated research outputs and fewer manual handoffs, freeing subject matter experts for higher-value work.
“The prompts are becoming complex… The collaboration has evolved to a different level.”
Kumar’s approach shows how enterprises convert AI tools into operational capability. Next steps include expanding agent thinking and standardizing prompt templates across functions. Teams should measure reduction in interruptions and time-to-decision as primary success metrics.

Key points from the article:

  • Prompts evolved from one-liners to detailed, multi-paragraph instructions.
  • Copilot Researcher replaced ad hoc colleague research and slide prep.
  • Copilot Studio enabled low-code agent creation for domain users.
  • Agents match talent availability and skills to projects via ERP data.
  • Integrated Copilot reduced context switching across Teams and Outlook.

    • Related Coverage:

    From the Source