How NATL Cut Quote Turnaround from 2–3 Weeks to 5 Days With CogniSure’s AI Underwriting Workbench
NATL automated Loss Run processing — reducing quote cycles to under 5 days, at over 99% accuracy (with human in the loop). Now they're automating the entire submission intake. Their underwriters can finally focus on what they were hired to do: profitable risk evaluation and client relationships.
A leading specialty insurer in commercial transportation.
National Interstate Insurance Company (NATL) is one of the leading specialty insurers in commercial transportation in the United States — serving trucking fleets, alternative risk programs, and specialty vehicle markets. NATL manages thousands of complex submissions annually, each requiring deep review of Loss Run documents across an enormous range of carrier formats.
Submission volumes have surged over the past five years. To stay competitive, we needed to move faster — from submission to quote — without simply adding more people to manually process every document. National Interstate has growth targets to hit year-over-year, and efficiency is how we get there.
A Manual Process That Struggled to Handle Scale and Complexity
Before CogniSure, every submission triggered the same manual chain. Loss Run documents arrived in hundreds of unstructured carrier formats. Teams reviewed each one manually, extracted the relevant data, and keyed it into customized Excel templates to feed the pricing model. Documents running into hundreds of pages per submission made the process highly complex — and volume was growing faster than headcount could.
Exploding submission volumes, with 2,500+ distinct carrier formats processed manually — resulting in slower quote cycles.
Submission Explosion
Submission volumes surged over the past five years — growth that outpaced what manual headcount could absorb.
Format Chaos
Loss Runs arrived in 2,500+ distinct carrier layouts — different fields, labels, and conventions in each.
Manual Document Review
Submissions running into hundreds of pages, reviewed line-by-line and re-keyed into Excel templates.
Slow Quote Cycles
2–3 weeks from submission to quote — too slow for a market where agents allocate business to the fastest insurer.
- 1Submission arrives
- 2Manual format review (2,500+ layouts)
- 3Page-by-page data extraction
- 4Excel re-entry into templates
- 5Pricing model build (3–5 hrs)
- 6Quote sent — weeks later
- 1Submission arrives
- 2CogniSure AI engine (patented LLM)
- 3Structured output · 99%+ accuracy
- 4Auto-mapped to NATL rating sheet
- 5Pricing model build (~1 hr)
- 6Quote sent — within days
Built by the industry,
for the industry.
CogniSure's platform was built by insurance professionals who understand the nuances of commercial P&C underwriting like no other. The result: a patented, LLM-based extraction engine that processes unstructured documents at benchmark-setting accuracy — across every carrier format, every line of business.
AI-Native, Insurance-First
Purpose-built for the insurance industry — not a generic AI tool adapted for underwriting. Decades of industry knowledge baked in from day one.
Patented LLM Extraction
Patented multi-modal LLM models process Loss Runs, ACORDs, schedules and MVRs with 99%+ accuracy across 2,500+ carrier formats. No training required.
★ PatentedAgentic Workflows
Carriers create their own workflows, generate reports their way, and customize on the go — without technical dependency. Business users stay in control.
Battle-Tested at Scale
Trusted by top US carriers over years of production use. Benchmark-setting accuracy forged through thousands of the hardest edge cases in the market.
Loss Summary Dashboard
Auto & Excess Liability claims analysis, cross-carrier — structured and ready for underwriting decisions in seconds.
Deliberate. Collaborative. Expanding.
Throughout Phase I, NATL and CogniSure held weekly check-in calls to review outputs, identify gaps, and make targeted refinements. This structured cadence — combined with CogniSure's ability to act quickly on feedback — is what took accuracy from 85–90% at launch to 99%+ today.
Starting with Proof
NATL provided representative Loss Run samples. CogniSure produced outputs. The team audited accuracy and turnaround, expanding complexity and edge cases iteratively. Nothing moved to production until it was right.
Loss Run Automation: The Manual Step, Eliminated
Platform goes live. Outputs mapped directly to NATL's proprietary rating sheet. API integration end-to-end. Results follow quickly.
Accuracy: 85–90% → 99%+
Weekly feedback calls, targeted refinements, rapid iteration. Accuracy climbed from 85–90% at launch to 99%+ — and the outputs became fully trusted by underwriting teams.
Transportation 360: Full Submission Lifecycle
NATL expands into the full Underwriting Workbench — ACORD applications, vehicle schedules, MVR reports, IFTA fuel reports, telematics, FMCSA data — all in one flow. Target: end-to-end processing under a day.
Agentic Underwriting at Scale
Push toward fully agentic submission triage — autonomous risk scoring, exception routing, and cross-portfolio benchmarking that compounds every quote NATL writes.
Feedback as
a Feature
Throughout Phase I, NATL and CogniSure held weekly check-in calls to review outputs, identify gaps, and make targeted refinements. This structured cadence — combined with CogniSure’s ability to act quickly on feedback — is what took accuracy from 85–90% at launch to 99%+ today.
The feedback loop is not an implementation artifact. It remains a core part of how the partnership operates — and what makes it feel different from a vendor relationship.
“CogniSure was — and continues to be — highly responsive in implementing customer-specific changes. Refreshing to work with a partner that approaches enhancements with a quick-strike mindset.
“The customer support team has always been quick to assess, quick to customize, quick to implement and move into production. Extremely refreshing.
Win-Win Relationship! Better Broker Experience
Efficiency gains were expected. What NATL didn’t anticipate was how faster, more accurate processing would transform its relationships with brokers and agents in the commercial transportation market.
In commercial transportation insurance, speed of feedback is a competitive differentiator. Agents allocate their submissions to insurers who respond quickly and clearly. CogniSure’s platform is now part of how NATL earns and keeps that business.
“We’re able to now provide real-time or very quick feedback — highlighting things we can provide back to the broker in very short order so that they have an expectation of what we need to proceed, or why we are not moving forward with a certain risk. That is a huge benefit.”
What changed,
and how much.
Measured across thousands of Loss Runs in production since September 2024 — spanning 2,500+ carrier formats.
Operational scale
Thousands of Loss Runs processed since Sept 2024 — without growing headcount.
Faster broker feedback
Real-time responses to brokers — a competitive differentiator in transportation insurance.
Underwriter focus restored
Teams spend time on risk evaluation and client relationships, not data entry.
Transportation 360: automating the entire submission lifecycle.
Loss Run automation cracked the biggest bottleneck. Phase II goes further — a full-spectrum submission intelligence layer that processes every document in one orchestrated flow.
End-to-End Automation
ACORDs, Loss Runs, schedules, MVRs, IFTA, telematics and FMCSA data — orchestrated by a single AI layer.
Centralised Workbench
Every submission delivered into one workbench, integrated with CRM — consistent insights, scoring and key data points across teams.
Agentic Orchestration
Multi-layered AI routes, scores and prioritizes submissions automatically — enriched with verified third-party intelligence.
One platform.
Every submission input.
ACORDs, loss runs, fleet schedules, fuel reports, telematics, regulator data — all flowing into a single underwriting workbench.
When the Underwriting Workbench is live, the question won’t be ‘how fast can we process this’ — it will simply be ‘what’s the risk and is it worth quoting.’ That’s what underwriting is supposed to feel like.
Sai Raman
Founder & CEO, CogniSure AI
Transform underwriting operations with AI.
See how CogniSure can take your submission intake from days to hours — without losing the judgment your underwriters bring.