Engineering Blog · Post #8

FIFO Assignment Engine: Round-Robin Rotation, Team Workload Balancing, SLA Tracking, and Manager Dashboard

How InsightUW eliminated cherry-picking in a MedMal underwriting team — and cut SLA breaches from 42% to 6% with fair, transparent, algorithm-driven assignment.

The Problem

Medical Malpractice is the line of business most susceptible to cherry-picking. Here's why:

MedMal submissions vary enormously in complexity. A solo dermatologist renewal takes 30 minutes to quote. A 450-bed teaching hospital with 28 surgical specialties, a Level I trauma center, and 10 years of claims history takes 3 weeks. Both land in the same shared inbox.

When Underwriting Assistants and underwriters self-select from a shared queue, the pattern is predictable:

Solo practitioners and small clinics get grabbed immediately — easy premium, fast turnaround, good hit ratio
Large hospital systems age in the inbox — complex, time-consuming, higher risk of declination
The best underwriters cherry-pick the highest-premium accounts — leaving junior staff with the dregs
Workloads become wildly unbalanced — one UW has 8 active submissions, another has 22
SLAs breach on complex accounts — because nobody wants to pick them up voluntarily

Real data from a mid-size MedMal carrier (pre-InsightUW):

Submission Type	Avg Days in Queue Before Pickup	Avg Days to Quote
Solo physician renewal	0.3 days	2 days
Small clinic (5-10 physicians)	1.2 days	4 days
Mid-size group (20-50 physicians)	3.1 days	8 days
Large hospital system	6.8 days	18 days

Large hospital systems — the highest-premium accounts — sat in the queue nearly 7 days before anyone claimed them. That's 7 days of the broker wondering if you want their business.

The InsightUW FIFO Algorithm

InsightUW's FIFO (First-In, First-Out) assignment engine removes human choice from the equation. Submissions are assigned to the team member with the oldest last assigned at timestamp — period.

graph TD subgraph New["New Submission"] A["SUB-2026-MM-0034 Valley Health Medical Center 320-bed hospital, 28 surgeons"] end subgraph Algorithm["Fifo Algorithm"] B["Get MED team roster"] C["Filter: active members only (exclude PTO, on-leave)"] D["Sort by last assigned at ASC (oldest first)"] E["Select top member (longest since last assignment)"] end subgraph Team["Med Mal Team Rotation"] F["Dr. Sarah Patel Last: 9:15 AM · Active: 14"] G["Michael Torres Last: 9:08 AM · Active: 12"] H["Jennifer Liu Last: 8:55 AM · Active: 11"] I["David Washington Last: 8:42 AM · Active: 10 ✅ Next"] end subgraph Assign["Assignment"] J["Assign to David Washington"] K["Update last assigned at = NOW"] L["David moves to bottom of rotation"] M["Notify David via app bell"] end A --> B B --> C C --> D D --> E E --> I I --> J J --> K K --> L L --> M

The Algorithm in Detail

Key properties:
- Deterministic — given the same team state, the same member is always selected
- Fair — over time, every active member receives approximately equal assignments
- Self-balancing — faster workers complete submissions sooner, but they get assigned again sooner too
- Transparent — every assignment includes the reason ("oldest last_assigned_at in MED team")

The Scenario

Valley Health Medical Center is a 320-bed community hospital in Portland, OR. They have 28 employed surgeons across 6 specialties, a Level II trauma center, and a claims-made policy with a 2019 retroactive date. Their current carrier non-renewed due to frequency concerns (4 claims in the last 2 years).

This is exactly the kind of submission that ages in a manual queue — complex, high-risk, non-renewed.

FIFO Assignment: 0.8 Seconds

sequenceDiagram participant Email as Broker Email participant Parser as Email Parser participant Fifo as Fifo Engine participant David as David Washington participant Mgr as Lisa Chen (Manager) participant Dash as Manager Dashboard Email->>Parser: Med Mal submission Valley Health, 320-bed hospital Parser->>Parser: Parse: Med Mal, Hospital Prof Liability 28 surgeons, non-renewed Parser->>Fifo: Create SUB-2026-MM-0034 Request MED team assignment Fifo->>Fifo: MED roster: 4 active members Sort by last assigned at Fifo->>Fifo: David Washington: 8:42 AM (oldest — 33 min since last) Fifo->>David: 🔔 Assigned: Valley Health Medical Center — Med Mal Fifo->>Dash: Dashboard updated: David: 10 → 11 active David->>David: Opens work queue Valley Health at top Mgr->>Dash: Views team workload All members within 10-14 range

Total time from email receipt to David's notification: 0.8 seconds.

In the old world, Valley Health would have sat in the queue for 6-7 days because no one wanted to tackle a non-renewed 320-bed hospital. With FIFO, David gets it immediately — and he knows the entire team is carrying a similar load.

David's Work Queue After Assignment

Priority	Type	Insured	LOB	Complexity	SLA	Status
HIGH	NEW	Valley Health Medical Center	MedMal	Hospital (320-bed)	7 days	ON TRACK
MEDIUM	IN_REVIEW	Cascade Orthopedic Group	MedMal	Group (12 physicians)	3 days left	ON TRACK
MEDIUM	IN_REVIEW	Dr. Amanda Foster	MedMal	Solo (dermatology)	5 days left	ON TRACK
LOW	QUOTED	Pacific Coast OB/GYN	MedMal	Group (6 physicians)	Waiting broker	—

Valley Health appears at the top — highest priority (non-renewed account, large hospital). David can't ignore it.

Manager Dashboard: Real-Time Workload Visibility

Lisa Chen, the MedMal team manager, sees the following dashboard:

Team Workload View

graph LR subgraph Dashboard["Med Mal Team Dashboard — Lisa Chen, Manager"] subgraph Sarah["Dr. Sarah Patel"] S1["New: 2"] S2["In Review: 8"] S3["Quoted: 4"] S4["Total: 14"] S5["SLA Breaches: 0"] end subgraph Michael["Michael Torres"] M1["New: 1"] M2["In Review: 7"] M3["Quoted: 4"] M4["Total: 12"] M5["SLA Breaches: 1"] end subgraph Jennifer["Jennifer Liu"] J1["New: 2"] J2["In Review: 6"] J3["Quoted: 3"] J4["Total: 11"] J5["SLA Breaches: 0"] end subgraph David["David Washington"] D1["New: 2"] D2["In Review: 6"] D3["Quoted: 3"] D4["Total: 11"] D5["SLA Breaches: 0"] end end

Workload Summary Table

Analyst	New	In Review	Quoted	Bound (MTD)	Total Active	SLA On Track	SLA At Risk	SLA Breach	Avg Days to Quote
Dr. Sarah Patel	2	8	4	6	14	12	2	0	5.2 days
Michael Torres	1	7	4	8	12	10	1	1	4.8 days
Jennifer Liu	2	6	3	7	11	11	0	0	4.1 days
David Washington	2	6	3	5	11	10	1	0	5.5 days
Team Total	7	27	14	26	48	43	4	1	4.9 days

Workload ratio (max/min): 14/11 = 1.27:1 — well within the healthy range.

SLA Tracking

Each submission carries an SLA based on priority and complexity:

Submission Type	SLA Target	Breach Alert At
Solo physician (renewal)	3 days	Day 2
Small clinic (< 10 physicians)	5 days	Day 4
Mid-size group (10-50 physicians)	7 days	Day 5
Large hospital (> 50 beds)	10 days	Day 7
Large hospital (non-renewed)	7 days	Day 5
Rush	4 hours	Hour 2

The dashboard highlights SLA status with color coding:

Color	Meaning	Action
Green	ON TRACK — more than 2 days remaining	No action needed
Amber	AT RISK — within 1 day of breach	Manager alerted
Red	BREACH — past SLA target	Manager + VP alerted, escalation

Manager Actions

Lisa can take several actions from the dashboard:

Action	When to Use	What Happens
Reassign	Analyst overloaded or out sick	Submission moves to another team member, audit logged
Escalate	Complex account needs senior attention	Priority raised, VP notified
Pause SLA	Waiting on broker for information	SLA clock stops, resumes when info received
Override FIFO	Specialist needed for specific account	Manual assignment with documented reason

Get Team Workload

FIFO Rotation Status

Why MedMal Benefits Most from FIFO

MedMal has characteristics that make manual assignment particularly harmful:

graph TD subgraph Problem["Why Med Mal Cherry-Picking Is Worst"] A["Extreme complexity variance Solo MD (30 min) vs. 450-bed hospital (3 weeks)"] B["Non-renewals are toxic Nobody wants to underwrite a carrier's rejected risk"] C["Long-tail claims Mistakes surface 3-5 years later — low incentive for quality now"] D["Surgical specialty risk Neurosurgery vs. dermatology — 50x premium difference"] E["Regulatory sensitivity State DOI watches Med Mal declination patterns closely"] end subgraph Solution["How Fifo Fixes It"] F["Complexity-blind assignment The algorithm doesn't know it's a 450-bed hospital"] G["Non-renewals assigned instantly No option to skip — it's your turn"] H["Accountability by assignment Every submission has a name and a timestamp"] I["Balanced specialist exposure Everyone gets a mix of easy and complex accounts"] J["Defensible to regulators Algorithm-driven, no bias, fully auditable"] end A --> F B --> G C --> H D --> I E --> J

Assignment Fairness

Metric	Before (Manual)	After (FIFO)	Change
Workload ratio (max/min)	3.2:1	1.3:1	-59% imbalance
Assignments to top performer	34% of total	26% of total	Normalized
Assignments to junior staff	12% of total	24% of total	Fair share
Non-renewed accounts: avg pickup time	6.8 days	0.8 seconds	-99.9%
Large hospital: avg pickup time	5.2 days	0.8 seconds	-99.9%

SLA Performance

Metric	Before (Manual)	After (FIFO)	Change
Overall SLA breach rate	42%	6%	-86%
Solo physician SLA breach	5%	1%	-80%
Small clinic SLA breach	18%	4%	-78%
Mid-size group SLA breach	45%	8%	-82%
Large hospital SLA breach	78%	12%	-85%

Business Outcomes

Metric	Before (Manual)	After (FIFO)	Change
Avg days to first quote	8.5 days	4.9 days	-42%
Quote-to-bind ratio	22%	31%	+41%
Broker complaints (per quarter)	14	3	-79%
Premium retention rate	71%	84%	+18%
UW team satisfaction score	3.1/5	4.2/5	+35%

The Biggest Win: Large Hospital Accounts

Before FIFO, large hospital accounts had a 78% SLA breach rate because nobody picked them up. After FIFO:

Large Hospital Metric	Before	After
Avg days sitting in queue	6.8	0 (assigned instantly)
SLA breach rate	78%	12%
Avg days to quote	22 days	12 days
Quote-to-bind ratio	15%	28%
Premium per bound account	$485K avg	$512K avg (+6%)

These are the carrier's highest-premium accounts. Getting them quoted 10 days faster increased the bind rate by 87% — and the premium per account actually increased because the UW had more time for proper analysis instead of rushing at the last minute.

Edge Cases: How FIFO Handles Exceptions

Scenario	FIFO Behavior
Team member on PTO	Excluded from rotation. Assignments skip to next. When they return, they re-enter rotation with their original last_assigned_at (may get multiple assignments quickly to catch up).
New team member joins	Enters rotation with last_assigned_at = NULL. Gets the first available assignment (NULLs sort first).
All team members at capacity	Assignment still occurs (FIFO doesn't consider capacity). Manager gets workload alert. Can reassign or hire.
Rush submission	Assigned via FIFO to next member, but with RUSH priority. Manager also notified.
Specialist needed	Manager overrides FIFO with manual assignment. Override is audit-logged with reason.
Duplicate submission detected	Not assigned. Flagged for manual merge.

InsightUW's FIFO assignment engine replaces inbox cherry-picking with algorithmic fairness — every submission assigned, every team member balanced, every SLA tracked. Request a demo to see the manager dashboard in action.