Replacing Excel-Based Referral Tracking with a Salesforce-Native Workflow Platform for a Specialty Home Pharmacy
HomeFree Pharmacy Services ran every stage of its patient referral lifecycle on separate spreadsheets. We built the Salesforce workflows, queue management, and Azure data pipeline that replaced them.
HomeFree Pharmacy Services coordinated every stage of its patient referral lifecycle from enrollment through insurance clearance to production and delivery across separate Excel files with no shared system of record, no live status between teams, and no audit trail when a handoff was missed.
Ideas2IT built Salesforce-native workflows that replaced the spreadsheet layer: admission objects with attached referral, patient, clinician, and Rx data; queues for each workflow stage; and task tracking across all teams. An Azure Cloud Functions pipeline connected Salesforce to Pioneer, preprocessing patient records in five days.
Pioneer data was preprocessed, deduplicated, and standardised in five days, with zero missing patient records after pipeline cutover. Admissions moved off Excel into a Salesforce-native queue with live visibility for all teams.
Phase 01
Building an entity model where every client's rules stay separate
Salesforce Workflow Platform: Replacing Spreadsheet-Tracked Admissions with a Single Orchestrated Referral Lifecycle
The constraint the Admissions team was working under was structural: every stage of the patient referral lifecycle had its own spreadsheet, and no spreadsheet knew what state any other was in. The first engineering decision was to model the full lifecycle inside Salesforce rather than alongside it.
Ideas2IT built admission objects with attached referral, patient, clinician, forms, Rx, and additional information. Queues were created for each workflow stage so the Admissions team could manage workload rather than chase status across files. Reports and analytics dashboards, built in the style of existing MedRec dashboards, surfaced admissions data on the Salesforce home page.
Phase 02
Connecting Salesforce to Pioneer with zero data loss
Azure Data Pipeline: Connecting Salesforce to Pioneer with Zero Data Loss and a Unified Patient Identifier
The second problem sat at the boundary between Salesforce and Pioneer: a large pool of patient records that needed to move between systems, be accessible with a unique identifier in both, and arrive without duplicates, format inconsistencies, or data loss.
Ideas2IT built a hashing mechanism on Azure Cloud Functions that generated a unique ID from mandatory patient fields, then ran the Pioneer dataset through a preprocessing stage in five days: deduplication, date-of-birth standardisation, name format normalisation, and phone number normalisation. An Azure Pipeline triggered on Salesforce activity sends data to Pioneer with the Unique ID and returns a patient status response.
The Outcome
From Batch-File Reporting and Siloed Spreadsheets to a Live, Auditable Referral Platform
| Category | Metric | Description |
|---|---|---|
| Preprocessing | 5 days | Entire Pioneer patient dataset deduplicated, standardised, and extended with Unique IDs |
| Data integrity | 0 missing | No patient records lost or unmatched after Salesforce-to-Pioneer pipeline went live |
| Automation | 4+ stages | Enrollment, insurance clearance, production handoff, and delivery coordination moved into Salesforce queues |
The referral lifecycle improvements followed directly from replacing a spreadsheet-per-stage model with a single Salesforce object that all teams read from and write to. Data integrity across the Salesforce-to-Pioneer boundary was the result of building the preprocessing and hashing mechanism before any records moved, not as a remediation step after. A five-day preprocessing run with zero missing records is what happens when the deduplication and standardisation logic is built into the pipeline rather than assumed of the source data.
Tech Stack
Common Questions
What teams usually ask before greenlighting a migration like this.
Phase 01 (Salesforce workflow platform) and Phase 02 (Azure data pipeline) ran as parallel workstreams. The Pioneer dataset itself, dedup, standardisation, Unique ID assignment, was fully preprocessed in 5 days once the pipeline was built.
Yes, this engagement involved referral, clinician, and Rx-linked patient data end to end. Ideas2IT is ISO 27001 and SOC 2 Type II certified, with RBAC and audit trails built into every engagement.
That was the exact starting point here: every referral stage lived in its own file with no shared status. We modeled the full lifecycle inside Salesforce as a single object all teams read from and write to, rather than bolting automation onto the spreadsheets.
We don't guarantee it in the abstract, we build for it. On this engagement, the hashing and deduplication logic was built into the pipeline before any records moved, which is why zero patient records went missing after cutover.
This engagement ran with a team of 5. Senior, cross-functional engineers operate as an embedded pod inside your environment rather than a detached vendor team.
Have a referral, records, or legacy-data problem like HomeFree's?
Tell us what's breaking. A senior engineer or domain specialist replies within 24 hours, no discovery-call runaround.