# The build map: a design tool for AI systems inside a bank

Source: https://www.bankingnewsai.com/agentic-banking/build
Last updated: Sep 17, 2026

Ten questions produce a design brief: the pattern, the controls, the documents that apply, the evals, the human gates and the questions an examiner will ask. Vendor-neutral. Twelve standard briefs, one per use case, are listed below.

## The ten questions

- **Who is affected** — Who does the output reach? Options: Staff only (An employee reads or uses it; nothing reaches a customer unreviewed); Customers, via a person (A person decides or sends, using the output); Customers, directly (The system answers, decides or acts toward a customer)
- **Reversibility** — If it is wrong, can it be undone? Options: Fully (A draft, a summary, a suggestion); With cost (Rework, an apology, a correction notice); Not really (Money moved, a decision communicated, a filing made)
- **Stakes** — What does a wrong call cost? Options: Low (Time and mild embarrassment); Material (Customer harm, a loss, a complaint); High (A credit decision, a regulatory filing, a payment, a legal exposure)
- **Knowledge** — Where does the truth live? Options: Stable documents (Policies, product terms, regulations, manuals); Live system state (Balances, cases, orders, limits, today's positions); Both (Documents to reason with, systems to check against)
- **Verifiability** — How is a good output recognised? Options: By a rule or a test (A schema, a reconciliation, a known answer); By judgment (A reviewer or a judge model can score it against a rubric); Hard to tell (Only time or an expert shows whether it was right)
- **Steps** — Are the steps known before you start? Options: Known (The same sequence every time); Partly (A known frame with branches); Open-ended (The path depends on what is found along the way)
- **Jurisdiction** — Which supervisors are in the room? Options: United States (Federal agencies plus state law); European Union (ECB, EBA, national supervisors, the AI Act); United Kingdom (PRA and FCA); More than one (The strictest regime sets the floor)
- **Delivery route** — How will the model be reached? Options: Provider API (A direct contract with the model provider); Through a cloud platform (The bank's existing cloud provider hosts the model); Undecided (Both routes are open)

## Patterns

| Pattern | What it is | When |
| --- | --- | --- |
| Augmented call | One model call with retrieval, tools and a structured output, inside code you control. | The task is well defined and the output can be checked. Most bank use cases should start here. |
| Workflow | Several model calls orchestrated by your code, in a sequence or a graph you designed. | Your code lays out the steps, the cost of error is real, and you need to observe every step. |
| Bounded agent | The model plans its own path through a set of tools toward a goal, inside a turn budget and an action envelope. | The path cannot be fixed in advance and the stakes allow it. Autonomy is kept in check by limiting tools, turns and what an action may touch. |

## Workflow shapes

- **Prompt chaining** — Each step's output is the next step's input, with a programmatic check between them.
- **Routing** — Classify the input first, then send it to a specialised path: a knowledge answer, a transaction on a live system, or a person.
- **Parallelization** — Independent sub-tasks run at the same time, or several attempts run and are compared, then results are aggregated in code.
- **Orchestrator and workers** — One model decomposes the task and hands scoped sub-tasks to workers; the orchestrator resolves gaps and conflicts when results return.
- **Evaluator and optimizer** — One model produces, another checks against explicit criteria, and the loop repeats until the check passes or a person is called.

## Tiers

- **Tier 1: Tier 1: a person decides** — High stakes, hard to undo, customer-facing. The model prepares, drafts and checks; a person decides and acts, and the record shows it. Gates: A named accountable executive signs off the use case before build, and the inventory records the tier.; A person reviews every output before it reaches a customer or a system of record; the review is logged with the reviewer's identity.; Independent validation before launch and outcomes analysis after it, on the model-risk schedule for a material model.; Adverse outcomes to a customer carry the specific reasons the law requires, produced from the decision record, not reconstructed afterwards.; A kill switch and a rollback path, tested before launch.
- **Tier 2: Tier 2: act with approval** — Material but recoverable. The model may prepare and, within limits, act, with a person approving anything that leaves the bank or touches a customer. Gates: Per-action approval by a competent reviewer for anything customer-facing or irreversible; sampled review for the rest.; Validation proportionate to materiality, with monitoring for drift on inputs and outputs.; An escalation route to a person that the customer can reach in one step.; Monthly review of the evals and the exception log by the accountable owner.
- **Tier 3: Tier 3: act within an envelope** — Low stakes, reversible, internal. The model may act within a defined envelope, with sampling, monitoring and a kill switch instead of per-action review. Gates: An action envelope: which tools, which systems, what value, what volume, what the model may never do.; Sampled human review of outputs and a weekly look at the exception log.; Turn and cost budgets, logged, with an automatic stop when exceeded.; A kill switch that any owner can pull.

## Primitives

| Primitive | What it is | In a bank |
| --- | --- | --- |
| Tools | Functions the model can call, with a typed schema; your code executes them and returns the result. | Every touch of a bank system is a tool. The action gateway sits between the call and the system. |
| Connectors (MCP) | An open protocol for exposing tools, data and prompts to models through a standard server interface. | One governed server per system of record beats ad hoc integrations; entitlements live on the server side. |
| Structured outputs | The model returns data in a schema you define, so code can validate and route it. | Anything that feeds a system or a decision record is structured, never free text. |
| Retrieval | Relevant passages from a document set are placed in the model's context before it answers. | For stable knowledge only. Live state comes from tools. |
| Subagents | Separate model contexts given scoped sub-tasks by an orchestrator, which merges their results. | Each subagent gets its own identity and tool set; the orchestrator resolves conflicts, not the workers. |
| Hooks | Deterministic code that runs on events: before a tool call, after an output, on completion. | Where policy checks, logging and blocking live; they do not depend on the model agreeing. |
| Packaged instructions (skills) | Reusable folders of instructions, scripts and references the model loads when a task matches. | How a bank's own procedures become repeatable; version them like code. |
| Turn and cost budgets | Limits on how many steps and how much compute an agent may use before it must stop or ask. | The simplest autonomy control there is; supervisors will ask for it. |

## Evals

- **Code-based** — Deterministic checks in milliseconds: schema, required fields, reconciliation, forbidden content. When: Every output, every run. The floor for anything that touches a system.
- **Model-based** — A judge model scores outputs against a written rubric. When: Interpretive tasks: summaries, narratives, answers. Calibrate against human scores monthly.
- **Human review** — Experts score a stratified sample, or everything, by hand. When: High stakes or novel behaviour, where neither code nor a judge can be trusted alone.

## Routes to a model

- **Provider API** — A direct contract with the model provider. Watch: One third party to diligence; data residency and retention terms are yours to negotiate.
- **Through a cloud platform** — The bank's existing cloud provider hosts the model and the bank's data stays inside that boundary. Watch: Two third parties in the chain and a concentration question; often the faster route through security review.
- **Mixed** — Different routes for different tiers or regions. Watch: Exit plans per route; the same evals across both so a switch can be judged.

## Standard briefs

- [Credit scoring & underwriting](https://www.bankingnewsai.com/agentic-banking/build/credit-underwriting) — a system that assembles and checks a credit application, and explains a decision
- [Fair lending & discrimination](https://www.bankingnewsai.com/agentic-banking/build/fair-lending) — a system that tests models and decisions for disparate treatment and impact, and documents the results
- [AML / KYC](https://www.bankingnewsai.com/agentic-banking/build/aml-kyc) — a system that triages alerts, assembles cases and drafts investigations for AML and KYC teams
- [Fraud detection](https://www.bankingnewsai.com/agentic-banking/build/fraud) — a system that scores and triages suspected fraud, and helps investigators and customers respond
- [Customer-facing chatbots](https://www.bankingnewsai.com/agentic-banking/build/customer-chatbots) — a system that answers customers' questions and completes simple tasks in a chat or voice channel
- [Model risk management](https://www.bankingnewsai.com/agentic-banking/build/model-risk) — a system that documents, validates and monitors models, and drafts the artefacts examiners read
- [Generative & agentic AI](https://www.bankingnewsai.com/agentic-banking/build/generative-agentic-ai) — a system that runs a bounded agent on internal tasks: research, drafting, operations exceptions
- [Third-party & vendor AI](https://www.bankingnewsai.com/agentic-banking/build/third-party-vendors) — a system that buys or configures a vendor's AI system and runs it under the bank's controls
- [Cybersecurity](https://www.bankingnewsai.com/agentic-banking/build/cybersecurity) — a system that supports detection, triage and response in the security operations centre
- [Data & privacy](https://www.bankingnewsai.com/agentic-banking/build/data-privacy) — a system that classifies, governs and answers questions about personal and risk data
- [Trading & capital markets](https://www.bankingnewsai.com/agentic-banking/build/trading-markets) — a system that supports research, surveillance and client communication in markets businesses
- [AI governance (general)](https://www.bankingnewsai.com/agentic-banking/build/governance-general) — a system that runs the AI inventory, policy and reporting that the board and supervisors ask for
- [AI-generated code & coding agents](https://www.bankingnewsai.com/agentic-banking/build/ai-generated-code) — a system that writes, reviews and ships code with AI assistance under the bank's architecture rules

## FAQ

### Is this advice on what our bank must do?

No. It is a design tool that maps common architecture choices to what regulators and public engineering documentation say, with dates and sources. Your compliance, risk and legal teams own the conclusions.

### Why is the pattern taxonomy vendor-neutral?

The three patterns and the workflow shapes come from public engineering guidance and appear, under similar names, in the documentation of every major provider. The brief never names a vendor because the controls are the same whichever model is behind them.

### Where do the documents come from?

From the regulation tracker on this site: 19 authorities, 164 documents, each linked to the official text and updated as rules change. A brief cites only documents that exist there.

### How do I use the brief?

Copy it into your intake or architecture-review template, keep the links, and treat the examiner questions as the test of whether the design is finished.

---

Canonical page: https://www.bankingnewsai.com/agentic-banking/build
Part of [BankingNewsAI](https://www.bankingnewsai.com/) — a free daily brief on AI in banking, an AI regulation tracker (19 authorities, 166 documents) and AI-strategy profiles of the 100 largest US banks. Markdown versions of every reference page: append `.md` to the page URL; index at https://www.bankingnewsai.com/llms.txt.