Juhi Tiwari
Juhi is a Research Lead at Kore.ai, driving market intelligence, competitive insights, and analyst-facing strategy that strengthens the company’s industry positioning. She also works as a content strategist, creating deep technical content that translates complex AI concepts into clear, compelling narratives. Her work supports Kore.ai’s vision, credibility, and thought leadership across the AI landscape.
Articles by this author
Agentic AI for business automation: Scaling with autonomy
Agentic AI for business automation enables AI agents to plan, reason, remember context, use tools, and autonomously execute complex enterprise workflows. This article explores AI-powered automation, multi-agent collaboration, memory, tool use, planning, observability, governance, and human oversight, alongside use cases in banking, customer service, procurement, insurance, lending, HR, and IT. It explains how enterprises can move from isolated AI pilots to scalable, governed autonomous operations.
AI observability: Monitoring and governing AI agents
AI observability provides the visibility enterprises need to monitor, understand, and govern autonomous AI agents in production. This article explores agent tracing, reasoning visibility, execution lineage, performance monitoring, security, drift detection, governance, and auditability. It explains how full-stack observability connects cognitive and operational telemetry to make AI agents more transparent, reliable, compliant, and accountable. Kore.ai provides agent-level tracing, behavioral analytics, lifecycle monitoring, and runtime governance.
A practical guide to AI agent evaluation
AI agent evaluation measures whether AI agents are reliable, accurate, secure, compliant, and capable of delivering business outcomes in production. This article explores agent evaluation frameworks, technical performance, groundedness, tool-use accuracy, safety, compliance, observability, regression testing, and business KPIs. It explains why evaluation must be continuous across the AI agent lifecycle and how Kore.ai Evaluation Studio supports model and agent evaluation, real-time tracing, custom evaluators, and continuous testing.