A day with Carla Reyes, VP Risk Management at Riverstone Construction Group.

Carla's day starts at 6:30 AM Central at her kitchen counter in Highland Park, Dallas, on a tablet next to a half-brewed cup. The risk-portfolio dashboard opens to the all-lines overnight delta view. One hundred and forty active claims across all coverage lines. Workers' comp 92 open ($9.4M reserves; the largest line). General liability 21 open ($3.6M reserves; one settled yesterday at $480K — that's the General Counsel call at noon). Commercial auto 17 open ($1.4M reserves; one subro pending). Builders risk + professional liability + the smaller lines add another 10 open claims at $0.4M reserves. Aged-claim watchlist (open >24 months): 11 — eight in litigation, three awaiting medical resolution. One new claim overnight: a workers' comp filing from the Houston Bayou Heights project site. **Severity-flagged.**

Three years ago this scan was a workers' comp loss-run from the carrier (two days late), a GL claims log in a different broker portal, a commercial auto report Carla had to ask for, and a manual reconciliation of which OSHA 300 entries matched which workers' comp filings (the answer was always "most but not all" and the reconciliation took half a day per quarter). Today it's one screen + a forty-second read. The Houston claim and the aged-claim watchlist are her morning priorities. Everything else is in a state the dashboard captures cleanly — reserves trending right, settlement velocity nominal, subro pipeline progressing, captive performance on plan.

Carla opens the Houston claim. Yesterday afternoon at 16:45 CT a laborer on the Bayou Heights project's east tower fell from an elevated work platform — a fall of approximately fourteen feet. ER-treated: distal radius fracture (orthopedic surgery pending), suspected mild concussion (CT negative for intracranial bleed, overnight observation, discharged 06:00 this morning). The OSHA 300 entry was filed by the project safety team at 17:20 yesterday; the first-notice-of-loss flowed into the WC carrier portal automatically thirty minutes later; the case appeared on Carla's dashboard as severity-flagged because of the head-injury element + the fall-from-elevation mechanism (both are predictive of escalation under Riverstone's claim-pattern modeling).

Carla sets initial reserves: medical $85K (orthopedic surgery + concussion follow-up + 12-week physical therapy estimate at standard Texas-jurisdiction multipliers); indemnity $45K (12-week TTD estimate at the operator's average weekly wage of $1,180; PPD waits for MMI); expense $15K (claim handling + an independent medical examiner if the case progresses + a litigation budget reserve given the head-injury risk). Total initial reserve: $145K. The litigation-hold flag activates pre-emptively. The notifications flow to the senior claims analyst, the Bayou Heights project superintendent, the Houston region safety manager, and the General Counsel. The case will get a senior claims analyst owner by 9 AM and a return-to-work workflow opened by EOB tomorrow. Three years ago this initial setup took 90 minutes across four systems + an emailed paperwork chain that lost the OSHA-300-to-WC linkage 30% of the time. Today it's 18 minutes + the linkage is built-in.

At 8:30 AM Carla joins the daily claims standup with the two corporate claims analysts and the WC clerk. Thirty minutes; same agenda every day. New claims filed in the prior 24 hours (today: four — the Houston WC severity case, a Dallas commercial-auto rear-end at an intersection, an Albuquerque GL third-party slip-and-fall on a jobsite, a minor Tulsa WC laceration). Claims closed (six today — four WC closing out at full return-to-work and maximum medical improvement, two GL closing — one the $480K settled premises-liability case, one a dismissal at summary judgment). Reserve changes greater than $25K (two today — a 2025 back injury progressing to surgery needs +$60K; a 2024 GL case where plaintiff counsel signaled litigation track needs +$120K). Fraud watch (one today — a Phoenix subcontractor WC claim filed Monday matched an ISO ClaimSearch hit against a 2023 claim under a different name + DOB delta of two days).

The senior claims analyst takes ownership of the Houston severity case + the GL reserve increase. The other analyst gets the Albuquerque GL + the Phoenix fraud flag (investigator assignment by end of day). The WC clerk handles the Dallas commercial-auto + the Tulsa minor WC + the four closures (closure paperwork + final-reserve trueup + the carrier-system reconciliation). The standup closes at 09:02. Three years ago this standup was forty-five minutes that started by trying to reconcile two different claim systems against each other; the meeting now runs on the unified dashboard view that doesn't need reconciliation because the data is the same record.

At 9:30 AM Carla joins the CFO for the quarterly TCOR forecast meeting. The FY26 budget submission is due Friday; this meeting is the alignment session before the formal submission. Carla brings a two-scenario forecast — base case + downside. Total TCOR FY25 ran at $28.2M against $1.1B revenue (2.5% of revenue). FY26 base case projects $28.4M (2.4% of revenue, slightly improving on the strength of the EMR trajectory + a softening WC market). FY26 downside ($31.1M; 2.6%) assumes two things: the Houston Bayou Heights case escalates to a $400K+ ultimate (head-injury + fall-from-elevation risk), and the GL market hardening Riverstone is hearing about from its broker materializes into a 10-12% rate increase rather than the 5% base assumption.

The CFO walks through the multi-state EMR forecasts with Carla — the platform's EMR projection model uses Riverstone's actual leading-indicator trajectory (hazard reporting rate + safe-behaviour observation ratio + near-miss-to-incident ratio across the project network) rather than the industry tables. The blended EMR is projected to move from 0.89 to 0.87 over FY26. The CFO asks the question he asks every quarter: what's the highest-probability single event that would invalidate the base case? Carla's answer (rehearsed because it's the same question every quarter): a sub-default exposure on a large GL claim — Riverstone's heavy-civil work has third-party exposure on roughly 30% of active projects, and a $5M+ event would punch through the umbrella attachment. The CFO acknowledges + asks for the FY26 budget submission Friday with both scenarios + a sensitivity table on the GL exposure assumption. Three years ago this forecast meeting required Carla to spend the prior week building the model from a workers' comp loss-run + a GL loss-run + a manual EMR calculation that didn't quite match the carrier's. Today the model is continuously current; the meeting is the conversation, not the data assembly.

At 10:45 AM Carla takes the call with the lead broker. Thirty minutes; same broker for the past six years; the relationship is the work and the call is the work. Today's agenda is Q4 renewal positioning. Workers' comp renewal lands January 1; Carla wants to push the SIR from $500K to $750K — the captive absorbs more first-dollar loss, premium drops accordingly, and the loss-development modeling she ran last month supports the move (the marginal $250K of SIR carries ~5 expected severe losses against captive surplus that comfortably absorbs them). Projected premium savings ~$420K. General liability renewal lands April 1; Carla wants to lead with the leading-indicator-to-incidents trajectory story — Riverstone's losses are running below market expectations and she wants an experience-credit reset against the hardening market. Commercial auto renewal lands March 1; the market is softening with new entrants who are using telematics + driver-scoring for risk selection; Carla wants to quote against three carriers plus the incumbent. The smaller lines (builders risk + EPLI + cyber + D&O) get handled by the line leads on staggered calendars + roll up at the Q4 board.

The broker confirms the market intel Carla expected. Construction GL is hardening 5-12% on catastrophe re-insurance pressure plus several large industry-wide losses. Workers' comp is stable with some softening in Texas + Arizona markets. Commercial auto is softening with telematics-driven new entrants. Cyber is stable. EPLI is hardening on retaliation-class trends. The broker pushes back gently on the WC SIR move — the market for $750K-attachment captive solutions has thinned somewhat — but agrees to take it to the WC underwriter. The call closes at 11:13 with a follow-up scheduled for ten days from now after the broker has talked to two WC markets + one GL market. Three years ago Carla walked into broker calls with whatever data her workers' comp loss-run gave her. Today she walks in with the actual EMR forecast, the leading-indicator trajectory, the loss-development model, and the captive performance summary — and the broker recognises it because Carla shares the summary with the broker the morning of the call as the starting point. The relationship works because both sides are reading the same data.

At 1:30 PM Carla works through the large-claim reserve adequacy review with the senior claims analyst. Twelve claims with reserves greater than $250K each; total reserves held $5.2M. Each claim gets a five-to-eight-minute review: current medical trajectory, return-to-work expectation, attorney involvement, settlement window, MSA (Medicare set-aside) status if Medicare-eligible, settlement-authority limits, anticipated ultimate.

Two reserve increases today. WC-2024-0419 — a back injury at the Tulsa project that's been progressing to surgery; MMI has been delayed twice; the medical trajectory now suggests a longer recovery + extended TTD; +$80K to $680K total reserve. GL-2024-0089 — a 2024 premises-liability case where plaintiff counsel signaled litigation track last week; the +$120K to $520K reflects the litigation-budget reserve component. WC-2023-0307 (an injured worker now MMI with stable PPD) is ready for settlement negotiation; the senior claims analyst gets authority up to $580K for a settlement attempt this week. WC-2022-0612 (a Medicare-eligible claim) is awaiting MSA approval through CMS — six-month estimated turnaround; nothing to do but wait. The remaining eight claims are routine; reserves carry forward at current levels. Reserve adequacy ratio for the large-claim portfolio: 94%, comfortably inside the 90-105% target range. Net change today: +$200K across two claims. Three years ago this review was a quarterly exercise that took two days because the data lived in three systems; today it's a 45-minute Wednesday rhythm.

At 3:00 PM Carla runs an EMR what-if model for the CFO. The COO has been working an acquisition target — a smaller Sun Belt construction firm with ~$120M revenue and operations in Florida + Georgia + Alabama + Mississippi (states where Riverstone doesn't currently operate). The CFO wants to understand the multi-line insurance impact of the acquisition before the next executive working session on the deal.

The platform's EMR what-if tool runs the scenario: take Riverstone's current 6-state EMR portfolio (blended 0.89), add the target's 4-state EMR portfolio (blended ~1.03 — the target's safety record is weaker than Riverstone's), and project the combined EMR over three policy years. Year 1: blended EMR shifts to 0.93 — worse than current standalone, because the target's EMRs are higher and the combined-entity exposure base will be re-rated under the higher-EMR contributors. Year 3 (assuming Riverstone applies its safety-program playbook to the acquired operations + observes the kind of leading-indicator trajectory Riverstone has run for its own portfolio): blended EMR reverts to roughly 0.88. WC premium impact: year-1 +$1.7M run-rate; year-3 roughly neutral to current baseline. Net of three-year cumulative pre-and-post-improvement: about $2.4M of incremental WC premium across the three-year horizon vs the no-acquisition counterfactual. Carla emails the model to the CFO with the framing he'll want: this is a transitional cost during integration; the playbook applies; the trajectory bends. The deal model still needs operational synergies + the integration cost — Carla doesn't do that math — but the WC premium impact is now quantified at the same fidelity as the operational synergies. Three years ago this scenario would have been a guess.

At 4:00 PM Carla runs the weekly OSHA-300-to-WC-claims integration reconciliation. The question this run answers: does every OSHA 300 entry filed by the project safety teams have a matching workers' comp claim, and if not, why not? The platform's integration runs continuously, but the weekly reconciliation is Carla's check on the integration logic. Forty-seven OSHA 300 entries across the past seven days, twenty-two active projects, six states. Forty-four matched cleanly to a workers' comp claim. Three outliers — explainable.

Outlier one: a property-damage-only event at the Tulsa project (a forklift struck a Sub-K wall section; no injury; no WC claim expected; the entry is on the 300 log as a "near-miss + property damage" classification per Riverstone's internal augmentation of the OSHA recordkeeping). Outlier two: a medical-treatment-only case at the Albuquerque project where the injured employee — after evaluation by the project occupational-health clinic — declined to file a workers' comp claim and elected to use his own group health coverage instead. Outlier three: a similar declination at the Dallas project. The two declinations are legitimate but need HR follow-up confirmation that the declination paperwork was completed per state-specific procedure (Texas and New Mexico have different declination-documentation requirements). Carla notes both for Friday's weekly HR-Risk sync. Zero unexplained gaps. Three years ago this reconciliation was a quarterly manual exercise: pull the workers' comp loss-run, pull the OSHA 300 log, cross-reference name + date + project + injury description in a spreadsheet, follow up on whatever didn't match. Sixteen hours of effort per quarter. Today the platform does the reconciliation continuously; Carla's role is reviewing the outliers + ensuring the declinations are properly documented. **The integration is the persona's headline value proposition.** Risk Managers without it spend half their time reconciling other people's data.

Carla closes the laptop at 5:30 PM Central. The Houston Bayou Heights case sits at the top of the dashboard with the initial reserve set + the litigation hold active + the senior claims analyst as owner; Friday's checkpoint with the treating physician's update will inform the next reserve action. The FY26 TCOR budget submission to the CFO is due Friday end-of-business with both scenarios + the GL exposure sensitivity table. The Q4 workers' comp renewal kickoff with the broker is next Monday — Carla will share the loss-development model + the $750K SIR positioning summary with the broker over the weekend. Tomorrow's calendar: the subrogation review on the $2.1M commercial auto claim (third-party-at-fault driver; the carrier's subro team is prepping recovery action). Friday brings the HR-Risk weekly sync with the two declination-documentation follow-ups + return-to-work review on four active cases. The subcontract-template update flows to the contracts team for Friday sign-off.

Tonight: dinner with her husband; a glance at the FY26 budget draft she'll refine tomorrow morning; a brief read of next Monday's broker call prep deck. Tomorrow brings a different mix — fewer claim-by-claim moments, more strategic writing. Carla has been doing Risk Management at construction operations for eighteen years; the difference between 2008 and today isn't the volume of claims (about the same) or the complexity (genuinely worse with multi-state expansions + cyber + EPLI lines). It's the visibility. She can see the portfolio, she can see the trajectory, she can see what's about to land, and the data that lives in the platform is the same data the CFO sees and the broker sees and the General Counsel sees. **The platform doesn't replace risk judgment. It replaces the reconciliation work that used to consume risk-judgment time.**