Skip to content
ale@portfolio:~/portfolio$ cat README.md

portfolio

Every case study here answers one question: measurable outcome.

third-party-integration.mdIPinfo · Nov 2025–May 2026

Stabilizing a Third-Party Integration

ShippedTeam: Eng + Infra + Sales + Web6 months
TL;DR

Stabilized a high-complaint, customer-facing third-party integration by fixing the existing version instead of rebuilding from scratch, validated through customer interviews, while migrating affected customers off a buggy version with zero broken workflows.

Outcome

  • ↓complaint volume, after systematic triage and prioritized fixes
  • ↑customer confidence, confirmed through post-fix interviews showing no major outstanding issues
  • →zero broken workflows during the version migration, old version kept live in parallel throughout
  • ↑sales enablement, via a dedicated roadmap reinforcing ongoing commitment to customers
Saw It
  • Inherited a third-party integration generating a high volume of customer complaints.
  • Open question was whether fixing the existing version could stabilize it, or if it needed a full rebuild from scratch.
↓
Figured It Out
  • Partnered with the lead developer to evaluate fix versus rebuild, and decided to fix.
  • Used Linear labels to categorize, list, and prioritize every reported complaint before touching code.
↓
Did It
  • Shipped fixes against the prioritized backlog, migrating affected customers off a buggy version with zero broken workflows.
  • Validated stability through engineering review and customer interviews before launch.

My Role

  • Partnered with the lead developer to evaluate fix versus rebuild for the integration.
  • Used Linear to label, prioritize, and track every customer-reported complaint, using AI to help prioritize and deduplicate overlapping issues and concerns into a clean roadmap.
  • Ran customer interviews to validate stability post-fix and surface any remaining issues.
  • Managed the safe migration of customers off a buggy version, keeping the old version live in parallel so no customer workflow broke.
  • Built a dedicated roadmap for Sales to share with existing customers, reinforcing commitment to continued improvement.
  • Investigated a direct integration path with the partner; determined it wasn't feasible.
  • Used AI to research competitor landscape, feature parity, and relevance across different customer use cases.
  • Proposed a full overhaul of the integration's dedicated website area, partnering with UX and using AI to prototype UX concepts.

Engineering Collaboration

  • Partnered directly with the lead developer on the fix-versus-rebuild decision and on prioritization.
  • Validated the prioritized roadmap together before committing to it, confirming the sequence of fixes would actually move the needle.
  • Worked through incremental fixes together, checking in on app stability as each one shipped rather than waiting for a single big release.
  • Coordinated the version migration to ensure no customer workflow broke during the fix rollout.
Full breakdown

Problem & Risk

Usability RiskFeasibility RiskBusiness Viability Risk
  • A third-party integration meant to give customers easy access to the data was generating a high volume of customer complaints.
  • Feasibility risk: unclear whether fixing the existing version could actually stabilize it, versus needing a full rebuild from scratch.
  • Business viability risk: needed to manage customer trust and expectations while a customer-facing integration was visibly unstable.

Customer Discovery & Validation

Evidence
  • Partnered with the lead developer in the integrations area to evaluate fix versus rebuild.
  • Used Linear labels to properly categorize, list, and prioritize every reported complaint.
  • Ran a round of customer interviews after stabilization work to confirm no major issues remained.
  • Investigated feature parity with competitor integrations.
  • Evaluated a direct integration path with the partner; confirmed it wasn't feasible.

Context

  • Inherited a third-party integration with a high volume of open customer complaints.
  • Faced a build-vs-fix decision early on, working directly with the lead developer to assess whether the existing codebase was salvageable.

Timeline

git log --oneline⏱ 6 months, Nov–May
triage: evaluated fix vs. rebuild with the lead developer, decided to fixNovember
label & prioritize: all complaints categorized and prioritized in LinearNov–Dec
version issue: small bug found in one version, affected customers contacted and migrated, old version kept live in parallelDec–Jan
stabilize: fixes shipped against the prioritized backlogNov–Mar
validate: engineering review plus customer interviews confirm stabilityMar–Apr
launch: green light given, roadmap shared with customers via SalesMay
Hand-drawn feature roadmap sketch grouping new features into columns for stability, security, data & control, and usability
roadmap.sketch: rough sketch, not an actual screenshot, new features grouped by stability, security, data & control, and usability

Who Benefits

  • Customers got a stabilized integration, clear migration support, and visibility into ongoing improvement plans.
  • Sales had a concrete roadmap to reinforce commitment and manage expectations with existing customers.
  • Engineering validated that a full rebuild wasn't necessary, avoiding a significantly larger investment.

Next Steps

  • Full UX overhaul of the integration's dedicated area on the website, proposed jointly with UX.