• contact@verticalserve.com
Home / Engineering / Post 59
Engineering Blog · Post #59

AI Research Assistant: Proactive Intelligence During the Underwriting Process

InsightUW's AI doesn't wait for questions — it proactively researches the insured, flags risks, analyzes market conditions, and suggests the right questions to ask.


The Problem

An underwriter receives a D&O submission for Solaris Energy Holdings. Before quoting, they need to research: Is there pending litigation? What's the financial trend? Any recent management changes? What are competitors offering? What questions should I ask the broker?

Today this research takes 45-60 minutes per submission — Googling the company, checking D&B, reading news, searching internal systems, asking colleagues. Most UWs skip the deep research on smaller accounts and hope for the best.

The InsightUW Approach

InsightUW's AI Research Assistant runs a comprehensive analysis with one click — aggregating company intelligence, risk flags, litigation records, market context, and LOB-specific questions into a structured research report. Plus a Q&A interface for ad-hoc follow-up questions.

graph TD subgraph Research["AI Research (One Click)"] Click["UW clicks 'Run AI Research'"] end subgraph Intel["6 Research Sections"] News["📰 Recent News<br/>sentiment analysis"] LIT["⚠ Litigation<br/>open/settled cases"] FIN["📊 Financials<br/>revenue, margin, debt"] Flags["🚩 Risk Flags<br/>specific concerns"] Market["📈 Market Context<br/>rates, capacity, trends"] QS["❓ Suggested Questions<br/>5 LOB-specific questions"] end subgraph Extra["Additional Context"] Peer["👥 Peer Comparison<br/>similar submissions"] QA["💬 Ask the AI<br/>free-text Q&A"] end Click --> News Click --> LIT Click --> FIN Click --> Flags Click --> Market Click --> QS Click --> Peer QS --> QA

Six Research Sections

Section What It Shows Example for Solaris Energy
Risk Narrative AI-generated assessment "Volatile revenue, pending ESG litigation, CFO departed"
Recent News Articles with sentiment [negative] "CFO departs amid restructuring" — Bloomberg
Litigation Cases with status/amounts Green Alliance v. Solaris (2026) — pending (ESG securities)
Risk Flags Specific concerns "Management instability", "Pending ESG securities litigation"
Market Context LOB conditions D&O: softening, -3% to -5%, ample capacity
Suggested Questions LOB-specific broker questions "Any SEC investigations?", "Board changes?", "D&B risk score?"

Q&A Interface

The research tab includes a free-text input where the UW can ask follow-up questions:

Question AI Answer
"What are the risk concerns?" "Risk flags: CFO departure, pending ESG litigation, publicly traded"
"What is the market rate trend?" "D&O: softening, -3% to -5%. Capacity ample."
"Any claims history?" "Green Alliance v. Solaris (2026) — pending ESG securities claim"
"What are the financials?" "Revenue volatile, operating margin 8.1%, debt ratio 0.52"

LOB-Specific Example

D&O for Solaris Energy Holdings:

UW Sarah Chen clicks "Run AI Research" on the AI Research tab. In seconds:

Risk Narrative: "ELEVATED RISK. Revenue volatile. Operating margin 8.1%. 2 negative news articles. 1 pending litigation (ESG securities). 3 risk flags: CFO departure, pending ESG litigation, publicly traded. Enhanced due diligence recommended."

News: Bloomberg reports CFO departure. S&P Global warns of rising ESG litigation for energy companies.

Risk Flags: Management instability. Pending ESG securities lawsuit. Publicly traded = heightened D&O exposure.

Suggested Questions: Sarah now has 5 specific questions to ask Lockton before quoting — including SEC investigations and board composition changes.

Sarah asks the AI: "Should we quote this risk?" AI responds: "Elevated risk profile due to pending litigation and management change. Recommend enhanced due diligence, higher retention, and senior UW review before quoting. If quoting, consider ESG exclusion and higher premium to reflect litigation risk."

What This Means for Underwriters

  1. Research in seconds — what used to take 45 minutes now takes one click
  2. Proactive risk flagging — AI surfaces concerns you might miss in manual research
  3. Market awareness — know whether the market is hardening or softening before setting terms
  4. Right questions — LOB-specific questions tailored to D&O vs Cyber vs GL vs WC vs Property
  5. Interactive — ask follow-up questions in natural language for deeper analysis

What's Next

Next: Three-Score Visibility — Priority, Quality, and Appetite on Every Submission


Want to see AI-powered research on your submissions? Request a demo.

See InsightUW run on your data

A 45-minute working session with a real broker email and your LOBs.

Request a demo