What We Built
Case Management
Orchestration Layer
Last-Mile Delivery
🔑 Environment & API Key Management
One of the less obvious but critical challenges was enabling safe development, testing, and debugging without impacting live delivery operations.
- Each customer needed separate dev/sandbox and production environments connected to Onfleet via scoped API keys
- Before and after go-live, we needed the ability to test changes, iterate on enhancements, and debug or do root cause analysis for integration issues — all without creating phantom deliveries or corrupting real operational data
- Spent significant time setting up the environment architecture and preparing both end-user admins and internal implementation teams to manage dev/prod integrations for each customer
- Documented the setup and key management process in the integration guide to make this repeatable
Unforeseen challenge: I underestimated how much time and attention environment management would require. In hindsight, I'd design the multi-environment architecture upfront rather than retrofitting it after realizing we needed safe iteration paths that wouldn't disrupt active meal delivery operations.
Nutritionist — Client Intake & Meal Plan (Mockup)
Case Manager — Service Record & Delivery History (Mockup)
Active Enrollments
| Program | Enrolled | Status | Frequency |
|---|---|---|---|
| Home-Delivered Meals — Medical | 01/15/2024 | Active | 3x/week |
| Wellness Check-in Calls | 01/15/2024 | Active | Weekly |
Recent Delivery History
| Date | Meals | Driver | Status | Notes |
|---|---|---|---|---|
| 04/23 | 2 | M. Torres (Staff) | Delivered | Left with front desk |
| 04/21 | 2 | M. Torres (Staff) | Delivered | Handed to client |
| 04/18 | 2 | B. Chen (Vol.) | Missed | Client not home |
| 04/16 | 2 | M. Torres (Staff) | Delivered | — |
Dispatcher — Kitchen Prep & Bag Labels (Mockup)
Meal Counts by Type
| Meal Type | Count |
|---|---|
| Regular | 12 |
| Diabetic | 8 |
| Low Sodium | 6 |
| Pureed | 3 |
| Renal | 4 |
| Gluten-free | 2 |
| Total | 35 |
Stop: 3 of 8
Dispatcher — Delivery Task Batch & Status (Mockup)
Delivery Tasks
| Client | Address | Diet | Meals | Driver | Status | |
|---|---|---|---|---|---|---|
| Margaret H. | 2841 S Highland Dr | Diabetic, Low-Na | 2 | Staff | Delivered | |
| Robert P. | 1190 E 3300 S | Regular | 1 | Staff | Delivered | |
| Dorothy M. | 4560 S 900 E, #12 | Pureed, GF | 3 | Volunteer | Manual Entry | |
| James W. | 6789 Cottonwood St | Renal | 2 | Staff | In Transit | |
| Helen K. | 3321 Murray Blvd | Diabetic | 1 | Volunteer | Missed | |
| Frank L. | 7200 Bengal Blvd | Low-Na, Soft | 2 | Staff | Ready |
Driver — Onfleet Mobile Experience
New volunteers could start delivering within minutes of receiving a text invite. Drivers marked deliveries as complete or missed (with a reason), and those events fired webhooks back to ClientTrack automatically.
Task List & Navigation
Optimized route with turn-by-turn nav. Delivery details, special instructions, and communication with dispatcher/recipient.
Easy to Learn, Easy to Use
Near-zero onboarding friction. This was the key differentiator vs. WorkWave — volunteers actually adopted it.
Driver ↔ Recipient Communication
Drivers could contact recipients directly — reducing missed deliveries when a client wasn't answering the door.
Dispatcher Dashboard — Route Assignment & Monitoring
After sending the batch from ClientTrack, planners used Onfleet's web dashboard to optimize driver assignments, sequence deliveries, and monitor routes in real-time.
Program Reporting — Delivery Outcomes (Mockup)
Program staff reported on delivery outcomes for program impact, funder requirements, and identifying clients with recurring misses.
Home-Delivered Meals — Monthly Summary
Clients with Recurring Missed Deliveries (3+ in period)
| Client | Missed | Total | Rate | Common Reason |
|---|---|---|---|---|
| Helen K. | 5 | 12 | 58% | Client not home |
| Arthur G. | 3 | 13 | 77% | Wrong address |
All mockups are representative. The actual ClientTrack UI used the platform's configurable form designer with API plug-ins.
⚠️ The Volunteer Driver Challenge
Onfleet's app dramatically improved driver adoption vs. WorkWave — but some elderly volunteers still preferred paper. We needed a dual-path approach.
Path A: Onfleet App (Happy Path)
Path B: Paper Route Sheet (Fallback)
Artifacts & Documentation
Build vs. Buy: Route optimization, driver apps, and real-time GPS are hard, specialized problems. Our value was connecting case management context — who needs what, with what dietary restrictions, and did they receive it — to the delivery operation. Onfleet's API and UX quality validated this decision repeatedly.