Program Management Portfolio
From Chaos
to Clarity
Technical project and operations leader with 7+ years delivering platform and systems initiatives across Product, Engineering, IT, and customer-facing teams.
Accomplished at translating ambiguous needs into executable plans, coordinating complex integrations, and improving operational readiness through clear ownership, risk management, and change adoption. Experienced with Gainsight, Zendesk, Jira, Confluence, and Asana.
24
SSO Implementations
35%
Response Time Reduction
95%+
Renewal Rate
Selected Case Studies
Click any card to read the full case study.
01
SSO Implementation Featured
Standardized the Single Sign-On Setup Process
Built a repeatable, cross-team SSO workflow that reduced implementation delays and duplicate account risk across 24 partner onboardings.
Context
SSO implementations required tight coordination between an institution’s IT team, internal partner stakeholders, Technical Support, and Engineering. For a while, the process worked — because one person held it all in their head.
What I Found
When the CTO who had built and owned the SSO process left, the process left with them. What had looked like a workflow was actually institutional memory. Rebuilding it fell to two people: me and an engineer. We didn’t just need to document what had existed — we needed to design something that could survive the next departure.
What I Did
- Reconstructed the SSO implementation workflow from scratch with an engineer, defining ownership and sequencing across all teams
- Standardized intake and scoping questions so prerequisites were established before technical work began
- Established a clear handoff model with defined timelines and accountabilities
- Documented the Unique Identifier as the primary source of truth to prevent duplicate accounts
- Published the full process as a centralized reference with a visual workflow map
Results
- Completed 24 SSO implementations using the rebuilt, standardized process
- Reduced risk of duplicate accounts and identity mismatches at the source
- Turned a process that had lived in one person’s head into something any team member could execute — and that could survive turnover
02
Support Operations
Zendesk Overhaul — 35% Response Time Reduction
Rebuilt a misconfigured support platform from scratch — and uncovered a hidden inbox holding tens of thousands of unanswered partner tickets.
Context
Zendesk had been set up incorrectly and underutilized. Macros were sparse, automations were untouched, and institutional knowledge about the tool lived in one person’s head. Partners were sending messages into a system that looked functional from the outside.
What I Found
While poking around the configuration — something I do when I want to understand a tool deeply — I discovered a secondary inbox that had been silently accumulating unread tickets for years. Nobody was monitoring it. Nobody knew. I flagged it to leadership and we built a cleanup plan together, escalating to VP level.
What I Did
- Audited the full Zendesk setup and identified root causes of the hidden inbox accumulation
- Rebuilt the macro library, replacing ad-hoc reply documents with consistent, reusable tooling
- Learned tagging and automation logic; implemented auto-close rules for tickets that didn’t require a reply
- Set up ongoing tagging to monitor misrouted messages and prevent future backlog buildup
- Documented the rebuilt workflow and trained colleagues on the new system
Results
- 35% reduction in time from message received to first reply
- Eliminated an unmonitored backlog that had gone undetected for years
- Gave CS leadership their first real visibility into support infrastructure — and demonstrated exactly what happens when tool knowledge lives in one person’s head
“
Keisha consistently found ways to improve processes and approached work as a true self-starter — balancing independence with strong collaboration across the team. Keisha’s strong analytical thinking and problem-solving skills made a meaningful impact, and any team would be fortunate to have Keisha as a colleague.Abbie Mui — Lead Product Designer and Researcher, Ladder Health
03
Process Automation
Self-Service SSO Attribute Mapping Tool
Supported rollout of a self-service tool that eliminated 96 manual emails per implementation cycle and cut configuration errors at the source.
Context
Every SSO implementation required 3–4 emails just to collect mapping details. In a year with 24 implementations, that was nearly 100 coordination touchpoints that didn’t need to exist. But the real cost wasn’t time — it was errors. Misconfigured mappings were landing back on Engineering as reversal requests, and the volume was becoming impossible to ignore.
What I Found
When I started tracking how often engineers were being pulled in to fix configuration mistakes that originated in email handoffs, the pattern was clear. The problem wasn’t that partners were careless — it was that asking people to map SAML attributes over email was an error-prone process by design. Moving that input to the partner’s own hands removed the translation layer where things broke.
What I Did
- Supported rollout of a self-service tool allowing IT users to map SAML attributes directly, without back-and-forth
- Documented the new workflow and built enablement resources for partner adoption
- Reinforced adoption through training and direct partner guidance during the transition
Results
- Eliminated 96 manual emails per implementation cycle
- Reduced configuration errors by moving input to the source — the partner’s IT team
- Freed Engineering from a category of reversal requests that had been accumulating quietly for months
04
Security & Compliance
Account Lockout Feature — TX-RAMP Compliance
Translated a compliance requirement into a shipped product feature, enabling the platform’s expansion into the Texas market.
Context
The business wanted to grow in Texas. Texas required TX-RAMP certification. TX-RAMP required a brute-force lockout control for non-SSO accounts. The feature wasn’t a product priority until it was a business blocker — and once it was, it needed to move fast.
What I Found
The compliance requirement didn’t arrive as a clean engineering spec. It arrived as a certification gap that needed to be translated into something buildable. That translation — from regulatory language to product logic, covering trigger rules, user notification, and account recovery — was the actual work.
What I Did
- Translated the TX-RAMP compliance requirement into a clear, actionable product spec for the lockout workflow
- Coordinated implementation details: trigger rules, user notification copy, and account recovery paths
- Aligned internal teams on rollout impact and messaging before launch
- Documented the full lockout lifecycle for support and partner-facing teams
Results
- Shipped a brute-force lockout system for non-SSO accounts
- Achieved TX-RAMP compliance requirements
- Unlocked the platform’s readiness for Texas market expansion — a business outcome that started with someone making sure a compliance gap didn’t get lost in translation
“
Their attention to detail, problem-solving skills, and collaborative nature materially improved the experience for MC’s partners — and their rise in the organization reflects the impact those skills and traits had on the team. Every organization needs a Keisha. Your organization needs a Keisha.Amma Marfo — Human-Centered Content Designer, Writer & Educator
05
Data Integrity
Validation Rules to Prevent Identity Collisions in Bulk Updates
Designed guardrails across four failure scenarios that reduced duplicate-user incidents and improved data integrity for customer-facing teams.
Context
Bulk uploads were creating duplicate users and incorrect contact attachments — and partners saw the mess directly in their dashboards. The problem wasn’t just operational friction; it was a trust issue with the product.
What I Found
Nobody asked me to find this. I kept seeing the same failure pattern and traced it upstream. The issue wasn’t that partners were doing something wrong — it was that the system had no guardrails to stop incorrect contact creation before it happened. The fix needed to live at the point of entry, not downstream in a correction workflow.
What I Did
- Defined a bulk action allowing admins to add contacts to existing users via CSV
- Established guardrails blocking incorrect contact creation across four distinct failure scenarios
- Coordinated rollout communication and usage guidance for affected internal teams
Results
- Reduced manual correction effort for duplicate-user and misattached contact scenarios
- Improved data integrity at the point of entry rather than downstream cleanup
- Gave customer-facing teams a more stable, predictable partner dashboard experience
06
Tooling & Visibility
SSO Login Sessions Chart
Partnered with Engineering to build a real-time SSO visibility tool that cut verification time and reduced back-and-forth for Support and IT teams.
Context
During SSO implementations, verifying that login was working correctly meant asking partners to test, waiting, and manually confirming identifier formats. There was no shared reference point — just emails and trust.
What I Found
The verification step was eating time at the end of every implementation — the moment where everything should feel done but didn’t. In a retrospective, someone said out loud what everyone had been working around: “we just need to be able to see this.” That was enough to define the requirement.
What I Did
- Captured requirements during an SSO implementation retrospective and aligned on success criteria with the build owner
- Partnered with Engineering to deliver a chart surfacing the five most recent SSO logins
- Defined the tool’s location in Admin and rolled out usage guidance for Implementation and Support teams
Results
- Gave Support and IT a shared, real-time reference for SSO testing and verification
- Reduced time spent validating identifier formats during active implementations
- Eliminated a category of back-and-forth that had previously slowed every implementation’s final verification step
07
User Experience
Session Timeout Extended to 60 Minutes
Turned a recurring complaint into a backend change — eliminating unexpected logouts during partner demos and training sessions.
Context
The platform’s 30-minute inactive session timeout was cutting off hour-long partner demos and training sessions mid-stream. It was a small setting with an outsized reputational impact: partners noticed, and internal teams were embarrassed.
What I Found
The feedback wasn’t coming through a formal channel — it was showing up repeatedly in passing, in prep conversations before demos, in post-session debrief. Nobody had formally documented it because it felt too small to escalate. It wasn’t.
What I Did
- Gathered recurring feedback from demo and training users and documented the business impact
- Built the case for the change and secured internal alignment
- Coordinated the backend update and communicated it clearly to affected teams
Results
- Extended session timeout to 60 minutes, eliminating mid-session logouts
- Improved the flow and professionalism of partner demos and trainings
- Demonstrated that small, unglamorous fixes with consistent documentation can move faster than larger requests — when someone takes the time to make the case
Let’s work
together
Looking to build systems that work for the people using them? I’d love to talk.