@borrowbetter/amlassdk

TypeScript SDK for the Alleviate MLAS platform. Wraps the GraphQL API for the lead service with automatic Azure AD authentication and full type safety.

Installation

npm install @borrowbetter/amlassdk

Requirements

• Node.js >= 18
• Azure AD client credentials (provided by your Alleviate account manager)

Quick Start

import { AlleviateMLAS } from "@borrowbetter/amlassdk";

const client = new AlleviateMLAS({
  environment: "sandbox", // or "production"
  auth: {
    clientId: process.env.AZURE_CLIENT_ID,
    clientSecret: process.env.AZURE_CLIENT_SECRET,
  },
});

Services

The lead service is accessed as a property on the client instance:

client.leadService

Lead Service

Manages lead creation, submission, status lookup, and referrals.

Create Lead

const result = await client.leadService.CreateLead({
  input: {
    firstName: "John",
    lastName: "Doe",
    email: "[email protected]",
    homePhone: "5551234567",
    address1: "123 Main St",
    city: "Atlanta",
    state: "GA",
    zipCode: "30301",
    estimatedDebt: 25000,
    loanPurposeId: 1,
    termsConsent: true,
  },
});

const { id, resultCode, note, bidValue } = result.createLead;

Submit Lead

After receiving a qualified result code (1013-1017) from createLead:

const submitResult = await client.leadService.SubmitLead({
  input: { leadId: id },
});

const { statusCode, crmRecordId, crm } = submitResult.submitLead;

Get Lead Status

const status = await client.leadService.GetLeadStatus({ id });

const { resultCode, updatedAt, bidValue } = status.getLeadStatus;

Create Referral

const referral = await client.leadService.CreateReferral({
  input: {
    firstName: "Jane",
    lastName: "Doe",
    email: "[email protected]",
    homePhone: "5559876543",
    state: "CA",
    dashboardId: 123,
    referralType: "ACCELERATOR_LOAN",
    payload: { customField: "value" },
  },
});

if (referral.createReferral.errors?.length) {
  console.error(referral.createReferral.errors);
} else {
  console.log(referral.createReferral.data);
}

Configuration

const client = new AlleviateMLAS({
  // Required
  environment: "sandbox" | "production",
  auth: {
    clientId: string,
    clientSecret: string,
  },

  // Optional
  headers?: Record<string, string>,  // Static headers added to every request
  timeout?: number,                  // Request timeout in ms (default: 30_000)
});

Custom Token Cache

By default tokens are cached in memory. For multi-process deployments (e.g. serverless, multiple workers) provide a shared cache implementation:

import { AlleviateMLAS, type TokenCache } from "@borrowbetter/amlassdk";

class RedisTokenCache implements TokenCache {
  async get() {
    return redis.get("mlas:token");
  }
  async set(token: string, expiresAt: Date) {
    const ttl = Math.max(0, Math.floor((expiresAt.getTime() - Date.now()) / 1000));
    await redis.set("mlas:token", token, { EX: ttl });
  }
}

const client = new AlleviateMLAS({
  environment: "production",
  auth: { clientId, clientSecret, cache: new RedisTokenCache() },
});

Raw Client Access

If you need direct access to the underlying graphql-request client:

client.rawLeadServiceClient

Type Exports

The service's generated types are available as a namespace export:

import {
  AlleviateMLAS,
  LeadService,
  type MLASConfig,
  type TokenCache,
} from "@borrowbetter/amlassdk";

Development

Setup

npm install
npm run dev      # codegen + tsup watch

Generated files (__generated__/sdk.ts) are excluded from source control and regenerated at build time.

Scripts

| Script | Description | |--------|-------------| | npm run build | Codegen + compile to dist/ | | npm run dev | Watch mode (codegen + tsup) | | npm run codegen | Regenerate TypeScript from GraphQL schemas | | npm run smoke | Run end-to-end smoke test against sandbox | | npm run lint | Biome lint check | | npm run format | Biome format + auto-fix | | npm run format:check | Biome CI check (no writes) | | npm run typecheck | TypeScript type check |

Smoke Test

cp .env.example .env.local  # add AZURE_CLIENT_ID and AZURE_CLIENT_SECRET
npm run smoke

Adding a New Service

1. Place the SDL schema at src/schemas/{service-name}/schema.graphql
2. Add operations under src/schemas/{service-name}/operations/
3. Add the service name to the SERVICES array in codegen.ts
4. Wire up the new client in AlleviateMLAS.ts
5. Export the namespace from src/index.ts
6. Run npm run codegen