AI Research Assistants

Introduction

Research assistants combine search + RAG + summarisation + citation. They power Perplexity, You, Glean.

Beginner analogy: Like a librarian + analyst — finds, reads, summarises, cites.

Understanding the topic

Core concepts to understand:

• Search → retrieve → summarise → cite → answer.
• Always show sources.
• Mark uncertainty.
• Cluster sources by viewpoint.
• Refresh stale results.

Syntax reference

Visual workflow / architecture:

Question → search/RAG → top K docs → grounded answer with citations

Real-world use

Perplexity hit $500M+ valuation in 18 months by nailing this pattern.

Best practices

• Always cite; reject sourceless answers.
• Cluster opposing views.
• Freshness signals matter.

Hands-on exercise

Interview preparation — practice these questions:

• Q1. Research assistant flow?
• Q2. Why citations matter?
• Q3. Cluster viewpoints — why?
• Q4. Freshness handling?