Driver-Hiring Pipeline Automation

What we're solving

Kreilkamp has open seats for 40 drivers. The supply isn't gone — qualified drivers are out there. The bottleneck is speed of response and after-hours coverage. Tenstreet's industry data¹ shows recruiters reach only 30–40% of applicants before they drift to another carrier, and 35–40% of driver applications come in nights and weekends when no one is at the desk.

Power BI already runs across Kreilkamp's businesses — Tim has noted the dashboards work well but "need attention all the time," meaning every signal lives in the data but every action still requires a human noticing. The pilot closes that gap for the hiring lane specifically.

The pilot is designed to change what Monday morning looks like for the recruiting desk. Every weekend application gets reached within five minutes. Every after-hours phone call gets a text back. The candidates who reply have answered ten qualifying questions before the recruiter's first sip of coffee.

The pilot does not replace the recruiter's judgment — it routes only judgment-ready conversations to him and gives Tim live visibility into what's happening in the pipeline.

What we build — three connected pieces, working inside Kreilkamp's existing iRecruit ATS

1. First-response automation (SMS)

SMS within five minutes of every new application, 24/7. Opens with what drivers decide on: pay (CPM + weekly estimate), home time, equipment age, route type. An AI agent carries the back-and-forth.

Why SMS not voice for the back-and-forth: drivers are mobile-native and reply between loads. Sustaining a 10-field conversation over voice loses candidates to fatigue mid-call. SMS is asynchronous — the driver answers during a fuel stop, conversation persists. Voice is reserved for after-hours intake (#3).

2. Pre-qualification capture — 10 FMCSA² fields, conducted over SMS

1. CDL Class A — valid, state, expiration
2. Endorsements held (H, N, T, X, P) — matched against Kreilkamp's fleet
3. CDL-A years of experience
4. Moving violations in last 36 months
5. Accidents in last 36 months (count + at-fault)
6. DUI / DWI history (5-year lookback)
7. Drug & Alcohol Clearinghouse status (driver self-report)
8. DOT medical card expiration
9. Employment gaps >30 days in last 3 years (per FMCSA §391.21)
10. Equipment-type preference + verifiable experience

The driver self-attests. The AI flags qualified for recruiter review vs. needs-clarification vs. hard-disqualified (CDL expired, DUI inside window). The AI does not make hire/no-hire decisions.

3. After-hours intake — voice entry, SMS continuation

A driver calls Kreilkamp's recruiting line after hours → AI receptionist asks name + number → immediate text-back opens the same SMS flow as #2. Kreilkamp doesn't lose the 35–40% of inquiries that come in nights and weekends.

Power BI — every interaction flows into the dashboards Kreilkamp already runs

Every event the pilot generates pushes to Kreilkamp's Power BI environment via Microsoft's standard push-dataset API. No new dashboard for the team to learn — the driver-hiring pipeline becomes visible inside the same operational layer Kreilkamp already uses across its businesses. Example alert built in week 1: "After-hours call unanswered for more than 10 minutes." From there, Tim and the recruiting team define which other thresholds matter — the dashboard moves from "needs attention all the time" toward "the crucial next action, delivered to the right person, promptly."

What Kreilkamp provides at kickoff

• iRecruit administrator contact at Cost Management Services for week-1 API discovery (the recruiting team currently uses iRecruit partially — discovery includes learning what is and isn't working today)
• A dedicated phone number for after-hours intake, or routing on an existing line
• 15 minutes with whoever maintains Kreilkamp's Power BI for data-feed setup
• Recruiter time: approximately 3 hours/week during weeks 1–2 (setup + calibration), then approximately 1 hour/week (qualifying-question tuning)
• Recruiter collaboration on script and AI-agent SMS-flow tuning. Any communication that represents Kreilkamp must fit the company's brand and ethos — only the recruiter can shape that. The AI executes language and judgment the recruiter has approved.

What gets confirmed in week 1 before building

• iRecruit API capability — webhook push vs. poll, and write-back permissions. If poll-only, OBB proceeds with 5-minute polling; if no write-back, candidate state lives in Power BI. Also: which iRecruit features are currently in use, which aren't, and where integration depth is appropriate.
• 10DLC carrier registration — federal SMS compliance³. Carrier approval takes 2–6 weeks; the first 4–6 weeks of the pilot are infrastructure setup before driver-facing traffic flows.
• Twilio Voice IVR⁴ + voicemail transcription accuracy on truck-cab background audio.
• Power BI push-dataset authentication handoff with Kreilkamp's BI analyst.

What OBB holds itself to (contracted metrics)

• Qualified-candidate conversations delivered to the recruiter — target ramp: 5 by day 45 / 30 by day 75 / 50 by day 90 (ramp starts after 10DLC clearance)
• Time from application to first SMS contact — median under 5 minutes once SMS is live
• After-hours phone calls captured and responded to — every call to the OBB line triggers SMS within 10 minutes, 24/7
• Recruiter-hours saved on initial qualifying calls vs. baseline (recruiter self-reports pre-pilot vs. post-pilot hours/week)

What gets observed but is not contracted: drivers signed (the recruiter's decision; depends on Kreilkamp's open roles and orientation cadence) and 90-day retention of pilot-funnel hires.

Out of scope (Phase 1)

• Pulling MVRs⁵, DAC reports⁶, or PSP records⁷
• Scheduling drug screens, DOT physicals, or orientation
• Making hiring recommendations of any kind
• Payroll, onboarding, or driver coaching
• Replacing iRecruit (it remains the system of record)
• Sourcing candidates (Kreilkamp's job boards and referral network remain in-house)
• Driver retention monitoring beyond what the data passively reveals