Let's Talk

Two People, Two Weeks: Replacing an Enterprise Tool with AI-Assisted Development

Two people, two weeks - how we replaced an enterprise tool with custom Salesforce functionality using AI-assisted development. Real results, real timeline.

Industry: Food Technology / CPG Company: Impossible Foods

The Situation

Impossible Foods needed to manage trade promotions—the pricing deals, distributor margins, and approval workflows that are critical for a CPG company selling through retail and foodservice channels. They’d been using an external tool called Blacksmith, but users hated it. Worse, it was producing incorrect margin calculations, which meant approvals were based on bad data.

The business rules lived across three separate spreadsheets maintained by different teams—product information, distributor details, and customer deal size thresholds that determined approval requirements and margin calculations. There was no dedicated development team available for the project.

The Challenge

  • Replace a disliked external tool with native Salesforce functionality
  • Consolidate 3 spreadsheets of business rules into a single system
  • Build complex margin calculations and approval logic
  • Deliver a production-ready solution with a two-person team
  • Complete the project in weeks, not months

The Approach

Working with Sarah Jamgotch from Impossible’s Sales Operations team, we used Salesforce’s AI-assisted development tools to rapidly translate business requirements into working code.

Sarah brought deep knowledge of the margin rules and user expectations. I brought Salesforce architecture expertise and knew how to structure prompts that would generate clean, maintainable code. Together, we iterated through requirements—translating spreadsheet logic into prompt instructions, reviewing AI-generated LWC interfaces and Apex controllers, and refining until we had production-ready functionality.

The AI handled boilerplate code and complex calculations quickly, while we focused on business logic validation and edge cases that required human judgment.

The Outcome

  • Production trade program management system delivered in approximately two weeks
  • Complex margin calculations now accurate and auditable
  • Three spreadsheets consolidated into native Salesforce
  • User adoption success—replacing a tool people actively avoided
  • Fraction of the cost of traditional custom development

Key Takeaway

AI-assisted development means sophisticated solutions are now affordable for growth-stage companies.

Why This Matters for Growth-Stage Companies

This project demonstrates a new category of what’s feasible for companies that need sophisticated functionality but can’t justify enterprise tool licensing or large development teams. The combination of domain expertise and AI-assisted coding opens doors that were previously cost-prohibitive.

AI-assisted development isn’t about replacing developers—it’s about what becomes possible when business expertise meets the right tools.