AYOGIImplement · Integrate · Automate

Portfolio · Salesforce

Salesforce, configured around how the business actually runs.

Sales Cloud and Service Cloud implementation for operations-led teams: data model, automation (Flows, validation, assignment logic), and the AI/document layer that connects Salesforce to the rest of a client's stack.

Case studies

Real, completed engagements. Company names and identifying details are anonymized to protect confidentiality; the figures are accurate.

Case Study · Veterinary Services Group

M&A data migration and a PII exposure catch

A multi-location veterinary services group growing through acquisition needed its fragmented M&A pipeline (three separate spreadsheets) consolidated into Salesforce, and a legacy debug trigger reviewed as part of a data audit.

3 spreadsheetsCompanies, deals, people
Dedup & matchBy phone/address, not just name
Live in Salesforce3,670 records via Bulk API
PII purged6,358 records, in the same audit
6,358
PII records purged after a forgotten debug trigger was found logging raw candidate data into production
3,670
Accounts, Contacts, and Opportunities migrated from three legacy spreadsheets, with 52+ duplicate collisions resolved
838+
Opportunities restored to visibility after fixing a hardcoded 500-row query cap that was silently hiding them
Case Study · Marketplace Platform

Contract and payment data integrity

A digital marketplace platform managing contractor and creator payment agreements had contract end-dates and payment schedules drifting from what the signed documents actually said, plus a metric that had been quietly diluted by duplicate records.

Signed contractThe actual source of truth
Drift foundSystem field didn't match the document
Verified & corrected78% of contracts, remainder flagged
78%
of contracts corrected with a verified end date sourced from the signed document, not an unreliable system field
6
contracts' payment schedules corrected to reconcile exactly with the signed total
1
duplicate-import bug found and fixed, correcting a per-unit cost metric that had been silently diluted
Case Study · Manufacturer, Service Cloud

Case escalation fix and a silent email failure

A manufacturer running Salesforce Service Cloud for customer support had reopened cases re-escalating incorrectly, and had no way to know that a core reply feature had a defect.

Case reopenedCustomer replies again
SLA fixedTimer now starts at reopen, not original creation
Escalates correctly80% were wrong before this fix
Agent repliesLooked successful in the UI
Email API never calledSince the feature launched
Customer receives itFixed and verified with a real test
80%
of reopened cases were re-escalating with the wrong SLA window before the fix (223 of 279 checked)
429
cases found stuck unclaimed in one queue for over 100 hours, surfacing a real operational bottleneck
1
silent failure caught: the "send reply" feature had never actually delivered agent replies to customers since launch
Case Study · Retail Technology Platform

A revenue rollup that was correct, until the data wasn't

A retail technology platform (RFID-based inventory tracking) needed a custom Sales Hub built to replace standard Salesforce layouts, plus ongoing platform maintenance.

Opportunity progressesAcross 7 pipeline stages
Rollup built & testedThe logic itself was correct
Data gap found0 of 42 had an Amount at all
0 / 42
Opportunities across 7 pipeline stages actually had a dollar amount populated, once the new revenue rollup was built correctly and tested

The rollup logic wasn't the problem. The underlying data was. Finding that distinction before shipping a fix is the point.

Where this covers Salesforce

Core

Sales Cloud

Pipeline, forecasting, and territory/assignment logic configured around how deals actually move.

Core

Service Cloud

Case management, SLAs, and escalation paths that hold up once support volume actually grows.

Marketing

Account Engagement

B2B marketing automation (formerly Pardot) sized for an SMB budget, not the enterprise B2C Marketing Cloud product.

Emerging

Agentforce

Salesforce's AI agent platform is still early for SMBs. Genuinely promising, but not yet a proven volume play, so it's worth scoping deliberately instead of defaulting into it.

Specialty: healthcare, payor/claims & benefits

Ash brings 10 years of hands-on experience in healthcare payor/claims, and a separate 10 years in benefits administration. That's why Health Cloud implementation and benefits-workflow automation is a genuine specialty here, not a generic add-on.

Concept Demo · Healthcare / Benefits

AI-assisted benefits plan intake

Benefits brokerages, PEOs, and self-insured employers running Salesforce routinely re-key plan data by hand from carrier-provided PDFs (coverage tiers, rates, eligibility rules), first into quoting, then again into a separate enrollment platform. Each re-entry point is a place for errors and lost hours to creep in.

Carrier PDFPlan document arrives
AI extractionTiers, rates, eligibility rules
Matches prior plan year → Salesforce record updated
Doesn't match → flagged for human review

It's a human-in-the-loop pattern, not a black box. A working, scaled-down prototype of this extraction pattern already exists. This is a capability demonstration, not a completed client engagement.

What's usually the actual constraint

For a 10–200 person team, the hard part of a Salesforce implementation is rarely the platform itself. It's making Salesforce reflect how the business actually operates, and keeping it from becoming an island. That's the same integration-layer problem that runs across every platform this practice works in: get the data model right, then make sure Salesforce isn't the fourth place someone has to manually re-type the same information.