Logistics is a data-intensive field where margins are tight and efficiency compounds. AI tools accelerate analysis, surfacing insights from freight invoices, carrier performance data, and route data that would take days to process manually.

1. Carrier Management and Freight Negotiation

RFP Development

Prompt: Help me build a transportation RFP for our parcel shipping lanes.

Our shipping profile:
- Annual volume: 850,000 packages/year
- Weight distribution: 60% under 5 lbs, 30% 5-20 lbs, 10% over 20 lbs
- Geographic distribution: 40% residential, 60% commercial
- Origin: 3 distribution centers (OH, TX, CA)
- Service requirements: 70% Ground, 20% 2-Day, 10% Next Day

Carrier performance priorities (ranked):
1. On-time delivery (98%+ target)
2. Damage rate (<0.3%)
3. Cost
4. Tracking visibility
5. Customer service responsiveness

RFP sections needed:
1. Company and shipping overview
2. Current volume and lane data (format for carriers to bid)
3. Service requirements and SLAs
4. Technology integration requirements (API, EDI)
5. Evaluation criteria and scoring
6. Timeline and process
7. Questions for carriers to answer

Also: List the key negotiating levers we should know about
(dim weight factors, fuel surcharges, accessorial schedules)

Freight Invoice Audit

Prompt: Analyze this freight invoice data for billing errors.

I have a CSV of freight invoices from the past quarter:
[Paste or describe invoice data: carrier, origin, destination, 
 weight, service level, billed amount, our expected rate]

Common billing errors to find:
1. Wrong service level billed (Ground billed as 2-Day)
2. Weight discrepancies (carrier's dimensional weight vs. actual)
3. Incorrect zone billing
4. Accessorial charges applied incorrectly (residential surcharge on commercial address)
5. Duplicate invoices
6. Rates not matching contracted tariff
7. Fuel surcharges exceeding contract percentage

For each error found:
- Invoice number and amount billed
- What it should have been (correct charge)
- Difference (amount to dispute)
- How confident I should be (clear error vs. borderline)

Total: Summarize total overbilling amount and top error types

Carrier Performance Scorecard

Prompt: Build a carrier performance scorecard from this data.

Data I have (last 6 months):
- On-time delivery rate by carrier (FedEx: 96.2%, UPS: 94.8%, USPS: 91.5%)
- Damage claims rate by carrier
- Invoice accuracy rate
- Average transit time vs. promised
- Customer escalations related to each carrier

Build a scorecard that:
1. Weights each metric by business importance (I'll tell you priorities)
2. Calculates overall performance score (0-100)
3. Identifies 3 specific improvement areas per carrier
4. Recommends volume allocation changes based on performance
5. Sets quarterly targets for each metric
6. Defines at what score level we should put carrier on improvement plan

Format: Carrier performance report suitable for QBR with carriers

2. Route Optimization and Fleet Operations

Fleet Route Planning

Prompt: Help me optimize this delivery route.

Context: Local distribution company, 1 truck, 18 stops

Stops with constraints:
- Stop 1: Restaurant (must deliver before 8am, large order)
- Stop 2-5: Office buildings (business hours only, 8am-5pm)
- Stop 6: Grocery store (receiving dock available 7am-2pm)
- Stop 7: Hospital (must deliver between 10am-12pm for specific department)
- Stop 8-12: Residential (any time, no access issues)
- Stop 13-15: Retail stores (receiving 9am-4pm)
- Stop 16: Cold storage facility (time-sensitive, temperature-controlled goods)
- Stop 17-18: Residential

Truck capacity: 26 feet, 16,000 lbs
Current load: 12,000 lbs
Start time: 6:00am at depot (address)
Drive time matrix: [provide or estimate]

Optimize for:
1. Earliest completion time
2. All time windows met
3. Minimum total miles
4. Fuel cost estimate

Output: Sequence list with estimated arrival and departure times

Driver Communication Templates

Prompt: Create driver communication templates for common logistics scenarios.

Scenario 1: Customer not available for delivery
Template for driver to send: brief, professional, with delivery options

Scenario 2: Address not found
Template with: confirmation of address, alternative options

Scenario 3: Delivery exception (damage noted at pickup)
Template with: damage description, photos taken, next steps

Scenario 4: End-of-day route not complete
Template to dispatch: stops remaining, reason (traffic/delay), ETA to complete

Scenario 5: Vehicle breakdown mid-route
Template to dispatch: location, vehicle status, stops affected

Format each template:
- Under 75 words
- Clear and professional
- Includes placeholder for specific details [brackets]
- Appropriate sense of urgency for each situation

3. Warehouse Operations

Slotting Optimization Analysis

Prompt: Help me analyze and improve our warehouse slotting.

Warehouse: 50,000 sq ft distribution center
Pick operations: Zone picking, 1,800 orders/day average
Current problem: High travel time per pick, pickers walking too far

Data available:
- SKU velocity data (how often each item is picked per month)
- Current bin locations for each SKU
- Order profiles (what items tend to be ordered together)
- Pick walk path (current vs. optimal zone)
- Aisle and location numbers

Analysis I need:
1. Identify top 20% SKUs by velocity (A-movers) — should be closest to pack area
2. Identify frequently co-picked items — should be slotted near each other
3. Flag B-movers currently in A-slots (inefficient use of prime locations)
4. Flag heavy items slotted at high pick locations (ergonomic risk)
5. Estimate travel time reduction from proposed changes

Slotting principles to apply:
- A-movers: waist-height, near shipping
- B-movers: mid-zone
- C-movers: back of warehouse
- Heavy items: floor level
- Co-picked items: adjacent locations

Output: Prioritized relocation plan (which SKUs to move first for biggest gain)

Receiving Process Improvement

Prompt: Design an optimized receiving process for our warehouse.

Current situation:
- Average receiving time: 4.2 hours per truck
- Errors: 3.2% of received units have data entry errors
- Bottleneck: Manual counting and paper-based recording
- Staff: 3 receivers per shift
- Volume: 8-12 trucks/day

Process I want to improve:
1. Appointment scheduling
2. Dock door assignment
3. Unloading
4. Counting and verification
5. Putaway
6. ASN (advance ship notice) reconciliation

For each process step:
1. Current state description
2. AI/technology recommendation to improve it
3. Expected time/accuracy improvement
4. Implementation complexity (easy/medium/hard)
5. Estimated cost

Also: What KPIs should I track to measure receiving performance?
Target: Reduce receiving time to 2.5 hours per truck

4. Last-Mile Delivery

Failed Delivery Analysis

Prompt: Analyze our failed delivery data and recommend improvements.

Delivery data (last 30 days):
- Total deliveries attempted: 12,450
- First-attempt success rate: 87.3%
- Failed delivery reasons:
  - Customer not home: 48% of failures
  - Wrong address: 22% of failures
  - Access code/gate needed: 15% of failures
  - Customer refused delivery: 8% of failures
  - Other: 7%

Re-delivery cost: $8.50 per re-attempt
Total cost of failures: [Calculate]

For each failure reason:
1. Root cause analysis (why is this happening?)
2. Prevention strategy (what changes before first attempt)
3. Recovery strategy (best practice for this failure type)
4. Technology solutions (what tools address this)
5. Expected first-attempt success rate improvement

Overall: If we improve first-attempt success to 94%, calculate:
- Annual delivery volume projection
- Cost savings
- Customer satisfaction impact

Customer Delivery Notifications

Prompt: Write a sequence of customer delivery notification messages.

Order: Customer ordered electronics, delivery expected Thursday
Carrier: FedEx Ground

Notification sequence:
1. Order shipped (day of shipment, Monday)
2. Out for delivery (Thursday morning, 7am)
3. Delivered successfully (with photo)
4. Failed delivery attempt (customer not home)
5. Final notice (package held at facility, will return in 5 days)

For each message:
- SMS version (under 160 chars)
- Email subject line
- Email body (under 100 words)
- Include relevant tracking info and clear next steps
- Brand tone: professional and helpful, not corporate-speak

5. Freight Cost Analysis

Lane Analysis

Prompt: Analyze our freight spend and identify savings opportunities.

I'll provide 6 months of shipment data:
[Origin city, Destination city, Weight, Mode (TL/LTL/Parcel), 
 Carrier, Rate paid, Service level, Delivery performance]

Analysis needed:
1. Top 10 lanes by total spend (identify where money is going)
2. For each high-spend lane: is current mode the most cost-effective?
3. Mode conversion opportunities:
   - Parcel shipments that are large enough to consider LTL
   - LTL shipments that could consolidate into TL
   - LTL shipments that could be served cheaper by parcel
4. Carrier mix analysis — are we diversified appropriately?
5. Accessorial cost analysis — what's driving extras?

For each opportunity:
- Current cost
- Estimated savings
- What would need to change
- Risk or tradeoff

Output: Freight cost reduction roadmap with prioritized actions

Customs Documentation

Prompt: Help me complete this commercial invoice for international shipment.

Shipper: ABC Manufacturing Corp (Cincinnati, OH, USA)
Consignee: XYZ Distribution Ltd (Hamburg, Germany)
Ship date: February 15, 2026
Carrier: DHL Express
Incoterms: DAP Hamburg

Products:
- 50 units industrial pumps, model P-2000, HS code 8413.70.9000
- Unit value: $450 USD
- Total value: $22,500 USD
- Made in USA

Additional info needed:
- What information must appear on the commercial invoice?
- What supporting documents are typically required for this type of shipment (EU import)?
- Are there any likely customs issues we should prepare for?
- What's the estimated import duty rate for this HS code into Germany?
- EORI number requirement?

Output: Completed commercial invoice template + document checklist

AI in logistics is most impactful when applied to recurring analytical tasks — freight auditing, carrier scoring, and route analysis — that currently consume analyst time. The output quality depends heavily on data quality, so investing in clean master data is a prerequisite for good AI results.