chore: added llm sample

usecases/llm/.env.sample

@@ -0,0 +1,13 @@
+export OPENAI_API_KEY="xxxxxx" 
+export PINECONE_API_KEY="xxxxxx"
+export PINE_ENV="xxxxxx"
+export AWS_ACCESS_KEY="xxxxxx"
+export AWS_SECRET_KEY="xxxxxx"
+export AWS_BUCKET_NAME="xxxxxx"
+export AWS_REGION_S3="xxxxxx"
+export PG_DB="xxxxxx"
+export PG_HOST="xxxxxx"
+export PG_USER="xxxxxx"
+export PG_PASSWORD="xxxxxx"
+export DOZER_APP_ID="xxxxxx"
+export DOZER_TOKEN="xxxxxx"

usecases/llm/Dockerfile

@@ -0,0 +1,17 @@
+# Use the official Python 3.9 image as the base image
+FROM python:3.9
+
+# Set the working directory within the container
+WORKDIR /usr/src/app
+
+# Copy the requirements.txt file into the container
+COPY requirements.txt .
+
+# Install dependencies from requirements.txt
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Copy your application files into the container
+COPY . .
+
+EXPOSE 8000
+CMD ["python", "-m", "chainlit", "run", "app.py", "--port" ,"8000"]

usecases/llm/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2023 Mrunmay Shelar
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

usecases/llm/README.md

@@ -0,0 +1,12 @@
+# Hyper-personalized chatbots using LLMs, Dozer and Vector Databases for Banking and Finance
+
+In this example, we will be using Dozer along with Chainlit, Langchain and Pinecone. 
+
+### Pre-requisites
+
+1. Pinecone account
+2. OpenAI API key
+3. Python 3.9/ Docker
+
+### Setup
+

usecases/llm/app.py

@@ -0,0 +1,254 @@
+import chainlit as cl
+import os
+from langchain.prompts import  PromptTemplate
+from langchain.chains import LLMChain
+from langchain.embeddings import HuggingFaceEmbeddings
+from langchain.schema import HumanMessage
+from time import sleep
+## For Vector DB
+import pinecone
+import uuid
+from langchain.vectorstores import Pinecone
+from langchain.chains.conversation.memory import ConversationBufferWindowMemory
+from langchain.chains import RetrievalQA
+from langchain.agents import Tool
+#from langchain import PromptTemplate, HuggingFaceHub, LLMChain
+from langchain.embeddings import HuggingFaceEmbeddings
+from langchain.agents import initialize_agent
+from langchain.embeddings.openai import OpenAIEmbeddings
+from langchain.chat_models import ChatOpenAI
+from langchain.chains import RetrievalQAWithSourcesChain
+from langchain.prompts.chat import ChatPromptTemplate,HumanMessagePromptTemplate,SystemMessagePromptTemplate
+from chainlit import user_session
+import os
+from pydozer.api import ApiClient
+from langchain.agents import ConversationalChatAgent, AgentExecutor
+import re
+import json
+
+def show_json(obj):
+    display(json.loads(obj.model_dump_json()))
+
+
+def wait_on_run(run, thread):
+    while run.status == "queued" or run.status == "in_progress":
+        run = client.beta.threads.runs.retrieve(
+            thread_id=thread.id,
+            run_id=run.id,
+        )
+        sleep(0.5)
+    return run
+### Dozer
+from openai import OpenAI
+client = OpenAI()
+DOZER_CLOUD_HOST = "data.getdozer.io:443"
+
+def get_api_client(app_id=None, token=None):
+    return ApiClient("financial_profile", url=DOZER_CLOUD_HOST, app_id=app_id,
+                     secure=True, token=token) if app_id else ApiClient("financial_profile", url="localhost:80")
+
+
+customer_client = get_api_client(app_id=os.environ.get("DOZER_APP_ID"), token=os.environ.get("DOZER_TOKEN"))
+
+def getCustomerData(input):
+    data = customer_client.query({"$filter": {"id":input}})
+    rec = data.records[0].record
+    id = rec.values[0].int_value
+    name = rec.values[1].string_value
+    income = rec.values[2].int_value
+    age = rec.values[3].int_value
+    dependents = rec.values[4].int_value
+    address = rec.values[5].string_value
+    prob = rec.values[6].float_value
+    credit_amt = rec.values[7].int_value
+    repay_status = rec.values[8].float_value
+    util_ratio = rec.values[9].float_value
+
+
+    return [id,name,income,age,dependents, address, prob,credit_amt,repay_status,util_ratio ]
+
+
+
+#user_env = user_session.get(".env")
+OPEN_API_KEY = os.environ.get("OPENAI_API_KEY")
+PINECONE_API_KEY = os.environ.get("PINECONE_API_KEY")
+PINECONE_ENV = os.environ.get("PINE_ENV")
+ # platform.openai.com
+model_name = 'text-embedding-ada-002'
+
+
+
+# find ENV (cloud region) next to API key in console
+index_name = 'langchain-retrieval-agent'
+pinecone.init(
+    api_key=PINECONE_API_KEY,
+    environment=PINECONE_ENV
+)
+
+
+#embeddings = OpenAIEmbeddings()
+
+
+text_field = "Answer"
+index = pinecone.Index('langchain-retrieval-agent')
+
+file = client.files.create(
+  file=open("/home/mrunmay/Desktop/Dozer/dozer-samples/usecases/llm/data/BankFAQs.csv", "rb"),
+  purpose='assistants'
+)
+
+# switch back to normal index for langchain
+#index = pinecone.Index(index_name)
+
+
+def customerProfile(input):
+    data = getCustomerData(int(input))
+    id  = data[0]
+    name1 = data[1]
+    income = data[2]
+    age= data[3]
+    dependents= data[4]
+    address= data[5]
+    prob = data[6]
+    credit_amt= data[7]
+    repay_status= data[8]
+    util_ratio= data[9]
+
+    profile = {
+        "id": id,
+        "name": name1,
+        "age": age,
+        "income": income,
+        "dependents": dependents,
+        "address": address,
+        "prob": prob,
+        "credit_amt": credit_amt,
+        "repay_status": repay_status,
+        "util_ratio": util_ratio
+
+
+    }
+    return json.dumps(profile)
+
+tools = [
+    {
+        "type": "function",
+        "function": {
+            "name": "customerProfile",
+            "description": "Useful when you need customer data to decide eligibility for a particular credit card. Use to check the probability of default and available balance to use it for eligibility",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "id": {"type": "number"}
+
+        },
+        "required": ["id"]
+
+    }},
+
+
+    }, 
+    {
+        "type": "code_interpreter",
+    }
+]
+assistant = client.beta.assistants.create(
+    name="Bank Assistant",
+    instructions="You are a bank chat bot. Help the client with all their banking needs. Calculate if the client is eligible for a credit card or not depending upon their customer profile",
+    tools=tools,
+    model="gpt-4-1106-preview",
+    file_ids=[file.id]
+
+)
+
+
+def setName(userName):
+    global name
+    name = userName
+
+
+
+
+global res
+@cl.on_chat_start
+
+async def start():
+    intro = "Hi there, I am an assistant for Bank A. I am here to assist you with all your banking needs! Please enter your id:  "
+
+    res = await cl.AskUserMessage(content=intro,timeout=200,raise_on_timeout=True).send()
+    global id
+    id = int(res['content'])
+    greeting = f"Hi {res['content']}. What brings you here?"
+    await cl.Message(content=greeting).send()
+    setName(res['content'])
+    # global credit 
+    # credit = getCredit(int(res['content']))
+    global customerinfo
+    customerinfo = customerProfile(int(res['content']) )
+
+
+    #print(customerinfo)
+    thread = client.beta.threads.create()
+    cl.user_session.set("chain", thread)
+
+
+
+@cl.on_message
+async def main(text_rep: str):
+    thread = cl.user_session.get("chain") 
+
+    #print(text_rep.content)
+    #print(thread.id, thread)
+    message = client.beta.threads.messages.create(
+        thread_id=thread.id,
+        role="user",
+        content=f"{text_rep.content}"
+    )
+    run = client.beta.threads.runs.create(
+        thread_id=thread.id,
+        assistant_id=assistant.id,
+        model="gpt-4-1106-preview",
+        instructions=f"Please help the user according to his ID: {id} ",
+        tools = tools,
+
+    )  
+
+    wait_on_run(run, thread)
+    if run.status == "requires_action":
+
+        tool_id = run.required_action.submit_tool_outputs.tool_calls
+        run = client.beta.threads.runs.submit_tool_outputs(
+        thread_id=thread.id,
+        run_id=run.id,
+        tool_outputs=[
+            {
+            "tool_id": tool_id[0].id,
+            "output": customerinfo,
+        }
+        ]
+
+        )
+
+
+
+
+    wait_on_run(run, thread)
+
+
+
+    messages = client.beta.threads.messages.list(
+        thread_id=thread.id
+    )
+    obj_string = str(messages)
+    pattern = re.compile(
+    r"ThreadMessage\(id='(?P<id>[^']*)', assistant_id='(?P<assistant_id>[^']*)', content=\[MessageContentText\(text=Text\((annotations=\[.*?\], )?value=\"(?P<content>.*?)\"\), type='(?P<message_content_text_type>[^']*)'\)\], created_at=(?P<created_at>\d+), file_ids=\[(?P<file_ids>.*?)\], metadata=\{(?P<metadata>.*?)\}, object='(?P<object>[^']*)', role='(?P<role>[^']*)', run_id='(?P<run_id>[^']*)', thread_id='(?P<thread_id>[^']*)'\)"
+    )
+    matches = pattern.finditer(obj_string)
+    for match in matches:
+        text_rep = match.group('content').replace('\\n', '\n').replace('\\"', '"')
+
+
+    #print(content)
+    await cl.Message(
+        content=text_rep ,
+    ).send()

usecases/llm/chainlit.md

@@ -0,0 +1,31 @@
+# Welcome to GoldenGate Bank ! 🏦💵
+
+Hi there, Customer ! 👋 We're excited to have you on board. GoldenGate Bank has been serving customers since 1970 . With a commitment to excellence, we have grown into a reputable financial institution, providing a wide range of services to meet your banking needs.
+
+# Our Services 🎄🎄🎁
+
+At 'GoldenGate Bank', we offer a comprehensive range of banking services to cater to your financial requirements. Our services include:
+
+**Savings Accounts**: Choose from a variety of savings accounts with competitive interest rates and personalized features.   
+**Current Accounts**: Our current accounts are designed to support businesses and individuals with seamless transaction capabilities.  
+**Fixed Deposits**: Benefit from attractive interest rates and flexible tenure options with our fixed deposit accounts.
+
+# Credit Cards 💳💳
+
+Experience the convenience and benefits of GoldenGate Bridge credit cards, tailored to suit your lifestyle and spending habits. We offer the following credit cards:
+
+1. **Rewards Card** 📦🎁
+Earn reward points on every purchase.       
+Eligibility: Minimum monthly income of $X,XX and a credit score of [Credit Score Requirement].
+
+2. **Travel Card** ✈️🧳  
+Enjoy exclusive travel perks and offers.  
+Eligibility: Minimum monthly income of $X,XX and a credit score of [Credit Score Requirement].
+
+3. **Cashback Card** 💵💵  
+Get cashback on eligible purchases.  
+Eligibility: Minimum monthly income of $X,XX and a credit score of [Credit Score Requirement].
+
+4. **Premium Card** 💰💲  
+Access premium benefits, including airport lounge access and concierge services.  
+Eligibility: Minimum monthly income of $X,XXX and a credit score of [Credit Score Requirement].

usecases/llm/docker-compose.yml

@@ -0,0 +1,13 @@
+version: '4.25' 
+
+services:
+  app:
+    build:
+      context: .
+      dockerfile: Dockerfile
+    volumes:
+      - .:/usr/src/app
+    env_file:
+      - .env
+    ports:
+      - '8000:8000'