Ben Snyder
Chime logo

Chime

Designing and launching an enterprise platform that powered every Chime operations team.

Abstract cover of Enterprise apps deck
Significant attention went into all artifacts while building out the enterprise platform.

Enterprise platform

I was brought on to help architect and unify the Chime enterprise platform in order to help get the company to IPO readiness.

Chime logo surrounded by cards for support, risk, fraud, disputes, security, engineering, policies, experiments, employee, and collections teams.
Every operations team needed cohesive tooling, but each had grown their own stack.

Understanding a fragmented operations ecosystem

We started by mapping every Chime operations function and the bespoke tools they relied on. The exercise exposed redundant investments, inconsistent data flows, and an opportunity to centralize the member experience for agents and specialists alike.

Three Google Sheets listing enterprise tooling features, prioritization models, and vendor matrices.
Feature inventory, prioritization scoring, and vendor capability matrices guided early decisions.

Quantifying scope and sequencing

Partnering with product and operations leads, we catalogued the full backlog of tooling needs. RICE scoring, cost modeling, and vendor comparisons helped us determine which surfaces to build, buy, or retire as we stood up the new platform.

Diagram comparing no centralized team, centralized team, and centralized platform against experience and efficiency metrics.
Aligning leadership on the investment case for a centralized platform became the mandate.

Aligning leadership on the platform bet

Visual models translated the tradeoffs between ad-hoc tooling, a service team, and a true platform. By framing the impact on experience, maintainability, and time to delivery, we secured funding for the long-term investment Chime needed.

Large-scale schema map showing member, account, and product data relationships across internal and external systems.
We mapped the full data lineage so every tool could trust the same normalized member record.

Designing a shared data foundation

The new platform hinged on a canonical schema. Design paired with architecture to document data relationships, third-party dependencies, and lineage so engineers and analysts could build safely on consistent definitions.

EnterpriseKit design system showcase with component stats and UI primitives.
EnterpriseKit shipped as the internal design system that accelerated every tool we launched.

Shipping EnterpriseKit to unify internal UX

We built EnterpriseKit—a component library with accessibility baked in—to keep dozens of product teams shipping consistent interfaces. Six teams adopted it in the first release, with thousands of installs powering all internal tools by the end of the year.

Penny application showing a member overview with transactions and identification details.
Penny’s core member view unified the workflows of 5,000+ agents across operations.

Launching Penny for operations

Penny became the flagship application for member support. We streamlined navigation, surfaced the right context per role, and rolled it out to more than five thousand agents who now rely on it for daily operations.

Penny application showing dispute investigation dashboard with timeline and transaction insights.
Advanced dispute tooling brought data science signals directly into the investigator workflow.

Embedding intelligence in dispute resolution

For the disputes team, we layered machine learning risk indicators and investigative timelines into the workflow. Investigators could triage cases faster, compare signals, and document outcomes without bouncing between tools.

Penny loan servicing screen showing payoff progress and loan details.
The lending module translated complex repayment logic into an intuitive progress model.

Supporting new product lines

As Chime expanded into lending, we extended Penny with domain-specific servicing tools. Agents could assess payoff progress, answer regulatory questions, and take quick actions with confidence during the pilot program.

MXP experimentation interface highlighting experiment stages, variants, and cohort builder.
MXP enabled product teams to design, launch, and monitor experiments without engineering support.

Operationalizing experimentation

MXP turned experimentation into a self-serve capability. From cohort creation to metric selection, the platform let teams move quickly while honoring governance requirements across Chime’s internal users.

Penny and Zendesk integrated interface showing a direct deposit queue with actions.
Integrating Penny with Zendesk connected ticketing, case management, and member context end-to-end.

Closing the loop with case management

Our final milestone paired Penny with Zendesk to orchestrate complex case queues. By threading agent actions, ticket states, and core account data together, we delivered a unified experience for the pilots that handled Chime’s most sensitive workflows.