Sidekick

What is Sidekick?

Sidekick is a an Agent Augmentation platform. It works alongside existing platforms, and provides a family of tools aimed at RA agents and Vetting agents. The tools gather evidence, and then provide an Overlay on the existing platform’s pages to allow agents to review evidence. It shows the agent which fields would be injected into, should the agent decide to accept the evidence. When accepted, the evidence is ingested into the platform with a single click. If rejected, we record the reason, so we can improve.

Sidekick integrates with the existing Rapid platform. It does not replace it. It adds a layer of structured intelligence at order receipt.

Sidekick is NOT AI, the agent is always in the loop. The evidence sources are recorded and can be ingested with the evidence making it Audit friendly.

Sidekick does not require that staff adopt any fundamentally different processes or tools. It does not become the tool of record, all evidence is still stored in the existing platform.

Sidekick can be introduced (or removed) with minimal impact.

Sidekick measures its own performance with rich telemetry data, so the models on this page stop being illustrative and start being facts.

The platform has been designed to be highly extendable with pluggable connectors. If new Rapid API interfaces become available, we can simply switch the ingestion method, the tool does not break. If we need to add new forms or tools we can do that easily. If we want to extend the screen scraping tools, we have the core and diagnostic tools ready. If we want to add real time dashboards or power BI integration that’s straight forwards.

New Tools and features can be added without impacting the core Rapid platform.

Future features could include, Zendesk integration, Live Dashboards for metrics, Power BI integration, Registry screen scraping, contextual Help tool for agents to help reduce on-boarding times.

Benefits

Sidekick produces linked improvements to the LEI operation. They compound.

• Easy cases never wait for an agent. If we need to ask the customer, we aim to do so immediately, regardless of queue length or time of day.
• Freed capacity drains the backlog non-linearly. When a team is at or near capacity, removing a fraction of the inbound volume has an outsized effect on average wait time. Agents have headroom for the cases that need them.
• The expensive long tail shrinks. Hard cases — needing multiple customer contacts — are identified early before an agent touches them.
• One customer interaction replaces several. Sidekick composes a single, structured evidence request covering every gap it can identify, replacing email ping pong. First-pass completeness rises; round-trip count falls.

All four effects are modelled on the detailed models. Experiment with the parameters yourself.

Sidekick’s genesis..

Working with the the LEI plugin demonstrated just how powerful the use of online forms can be. We use them for gathering L2 information where needed. They are quick to produce, and can guide the customer step by step, using logic to hide unnecessary and confusing questions. We can prefill them and email the link. The users can pause and resume, and upload documents as needed. They proved to be very effective with high accuracy and completeness from the customers, much superior to complex email templated requests for information, especially as we can request the information automatically on order receipt. However, getting the data out of form tool was hard, without an API to ingest the form data into the Rapid platform, and in any case, it needs agent review before it can be used. The concept of DOM injection via a custom Browser plugin solves both of these issues. Here, we can open an order on the Rapid platform that has a completed L2 form, see and review the relevant data, then chose to accept or reject the evidence, with one click. The DOM injection adds the information as an agent would do on the page, but fast. It soon became obvious that this pattern of Evidence Collection and Ingestion (ECI) had multiple applications, for example for Trusts and Funds. Other tool’s were added too, to help with establishing Empowerment, like the Registry API look up and the LOA/POA form ECI component. Screen scraping has it’s place for some registries, so we have built that in, and have used it initially on our Website Association tool which searches for applicants on their website, ranks and screens hits and presents a screenshot and link of any it finds. Sidekick grew into a generic platform that could use it’s toolset to run many applications, so we have architected it to be pluggable and extendable from the start.

How it fits in

One of the key enablers is the use of a small browser extension that runs on the agent’s machine. It is aware of the when the agent is on the relevant page. It talks to the Sidekick server, and presents a small side panel on the page, on which the agent can view and accept or reject evidence and highlights on the page which fields a piece of evidence is targetted at, then it manages the DOM injection of evidence.

Everything else runs on the Sidekick server, which orchestrates Order awareness, Evidence collection workflows, logs, telemetry and tool usage. The Sidekick server can be self hosted on an internal server, or could be cloud based.

The server also provides an Admin panel that is used to configure the platform, manage users, and view orders, cases and logs. It will provide telemetry dashboards.

See where the time actually goes →

What Sidekick is actually doing

Sidekick produces four linked improvements to the LEI operation — each visible in the detailed models. Together they compound:

• It acts at order receipt, so easy cases never queue for a human at all.
• It returns capacity, which drains the backlog non-linearly when the team is under load.
• It collapses the expensive tail by resolving hard cases up front.
• It raises first-pass completeness, so a single customer interaction does the job instead of three.

None of these replaces the core system, the workflow engine, or human judgement. Sidekick stays quiet when it has nothing useful to add, and measures even its own silence. The benefit is speed, consistency, and the agent’s scarce attention spent where it genuinely matters.

These models are only as good as the data behind them. The next step is discussions with stakeholders and instrumenting the real workflow — so the figures on detailed models stop being illustrative and start being measured.

About Breeze Identity

Sidekick didn’t come from a whiteboard. It came from years inside CA vetting operations — running the queues, owning the performance numbers, and feeling first-hand where the hours actually go. I introduced CRM-based case management to bring order to that work, and I’ve sat on the audit side too, which is why evidence, traceability, and defensible process are built into Sidekick from the start rather than bolted on. Data driven metrics are foundational in making process improvements, so Sidekick measures everything. That operational grounding is paired with a genuine care for the people at both ends of a case: the applicant who deserves a clean, un confusing UX, and the agent whose scarce time shouldn’t be spent on what a machine could have prepared. I build to engineering best practice — properly architected, maintainable, honest about its own limits — and I have a real appreciation for what good tooling does for a team, because I’ve lived the difference between fighting with tools and being carried by them. Sidekick is the tool I wish I’d had.