feat(api): add gpt response caching

apps/api/prisma/migrations/20240716180407_add_cached_gpt_response/migration.sql

@@ -0,0 +1,10 @@
+-- CreateTable
+CREATE TABLE "CachedGptResponse" (
+    "id" TEXT NOT NULL,
+    "createdAt" TIMESTAMP(3) NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    "updatedAt" TIMESTAMP(3) NOT NULL,
+    "query" TEXT NOT NULL,
+    "response" TEXT NOT NULL,
+
+    CONSTRAINT "CachedGptResponse_pkey" PRIMARY KEY ("id")
+);

apps/api/prisma/schema.prisma

@@ -190,3 +190,12 @@ enum CategoryType {
   INCOME
   EXPENSE
 }
+
+model CachedGptResponse {
+  id        String   @id @default(cuid())
+  createdAt DateTime @default(now())
+  updatedAt DateTime @updatedAt
+
+  query    String
+  response String
+}

apps/api/v1/services/ai-cache.service.ts

@@ -0,0 +1,24 @@
+import prisma from '../../lib/prisma'
+
+export async function findAiCacheByQuery({
+  query,
+}: {
+  query: string
+}) {
+  return await prisma.cachedGptResponse.findFirst({
+    where: { query },
+    orderBy: { updatedAt: 'desc' },
+  })
+}
+
+export async function createAiCache({
+  query,
+  response,
+}: {
+  query: string
+  response: string
+}) {
+  return await prisma.cachedGptResponse.create({
+    data: { query, response },
+  })
+}

apps/api/v1/services/ai.service.ts

@@ -1,6 +1,8 @@
 import type { CategoryType } from '@prisma/client'
 import { OpenAI } from 'openai'
 import { getLogger } from '../../lib/log'
+import { createAiCache, findAiCacheByQuery } from './ai-cache.service'
+import { hashFile } from './file.service'
 
 const ASSISTANT_ID = process.env.OPENAI_ASSISTANT_ID!
 
@@ -54,6 +56,23 @@ export async function generateTransactionDataFromFile({
   }[]
 }) {
   const log = getLogger(`ai.service:${generateTransactionDataFromFile.name}`)
+  const additionalInstructions = `note must be in ${noteLanguage} language but do not translate names. Categories: ${JSON.stringify(categories?.map((cat) => ({ id: cat.id, name: cat.name, icon: cat.icon, type: cat.type })) || [])}`
+
+  // If cached, return cached response
+  const fileHash = await hashFile(inputFile)
+  const cacheQuery = `transaction_file:${JSON.stringify({
+    fileHash,
+    additionalInstructions,
+  })}`
+
+  log.debug('Checking cache for query: %s', cacheQuery)
+
+  const cachedResponse = await findAiCacheByQuery({ query: cacheQuery })
+
+  if (cachedResponse) {
+    log.info('Found cached response for query: %s', cacheQuery)
+    return JSON.parse(cachedResponse.response)
+  }
 
   const file = await uploadVisionFile({ file: inputFile })
 
@@ -71,8 +90,6 @@ export async function generateTransactionDataFromFile({
   log.info('Created thread with uploaded file. Thread ID: %s', thread.id)
   log.debug('Created thread with uploaded file. Thread details: %o', thread)
 
-  const additionalInstructions = `note must be in ${noteLanguage} language but do not translate names. Categories: ${JSON.stringify(categories?.map((cat) => ({ id: cat.id, name: cat.name, icon: cat.icon, type: cat.type })) || [])}`
-
   log.debug(
     'Running assistant on thread.\nAssistant ID: %s\nAdditional instructions: %s',
     ASSISTANT_ID,
@@ -106,6 +123,12 @@ export async function generateTransactionDataFromFile({
 
     cleanup()
 
+    // save to cache
+    createAiCache({
+      query: cacheQuery,
+      response: JSON.stringify(aiTransactionData),
+    })
+
     return aiTransactionData
   }
 

apps/api/v1/services/file.service.ts

@@ -0,0 +1,15 @@
+import { createHash } from 'crypto'
+import { getLogger } from '../../lib/log'
+
+export async function hashFile(file: File) {
+  const log = getLogger(`file.service:${hashFile.name}`)
+
+  log.debug('Hashing file. File size: %d', file.size)
+
+  const buffer = Buffer.from(await file.arrayBuffer())
+  const hash = createHash('sha256').update(buffer).digest('hex')
+
+  log.info('Hashed file. Hash: %s', hash)
+
+  return hash
+}