Beach Ready?

Sailing for An Early Retirement!

Option 1: Robo Advisor for Retirement Plans

A bot RoboAdvisor is built for retirement planning on Amazon Lex.

Demo:

Scenario 1: when the provided age falls between 0 and 65 with an investment of at least $5000

Scenario 2: when investing less than $5000 or the age provided falls outside the range of 0 to 65

Files

---

• Lambda Function

• Robot on Lex

• Demo Video

---

Details

Two more utterances are added to those provided in the instructions as follows:

• How to retire early and take ​{riskLevel}​ risk
• I need ​{investmentAmount}​ by ​{age}​ to retire

Recommendations are built under the recommend_portfolio(intent_request) function as follows:

    """
    Performs dialog management and fulfillment for recommending a portfolio.
    """
    risk_level = get_slots(intent_request)["riskLevel"]

    # define investment portfolio recommendations

"""
    Responses based on selected risk levels.
    """
    risk = {
        "None": "100% bonds (AGG), 0% equities (SPY)",
        "Very Low": "80% bonds (AGG), 20% equities (SPY)",
        "Low": "60% bonds (AGG), 40% equities (SPY)",
        "Medium": "40% bonds (AGG), 60% equities (SPY)",
        "High": "20% bonds (AGG), 80% equities (SPY)",
        "Very High": "0% bonds (AGG), 100% equities (SPY)"
    }

with initial recommendation

# Get the initial investment recommendation
    initial_recommendation = risk[risk_level]

Test Results on Lambda Function:

Results (click me):

Test Files (provided)

• correct_dialog.txt
• age_error.txt
• incorrect_amount_error.txt
• negative_age_error.txt

Customize riskLevel Slot

Prompt cards on risk levels (click me):

Design on RecommendPortfolio Intent:

Overview on Amazon Lex (click me):

Next Steps

• Construct post-confirmation intents using get_slots on riskLevel responses. In the current version, the Confirmation Prompt box was unchecked on Amazon Lex intent builder as it "short-circuit" the trigger to select riskLevel.
  • If ordering becomes an issue, define a separate function recommended_portfolios(risk) shown below could be an alternative to the current risk dictionary defined under the function recommend_portfolio(intent_request):

Click me for details on an alternative function:

 # define investment portfolio recommendations
def recommended_portfolios(risk):
"""
Responses based on selected risk levels.
"""
risk = {
    "None": "100% bonds (AGG), 0% equities (SPY)",
    "Very Low": "80% bonds (AGG), 20% equities (SPY)",
    "Low": "60% bonds (AGG), 40% equities (SPY)",
    "Medium": "40% bonds (AGG), 60% equities (SPY)",
    "High": "20% bonds (AGG), 80% equities (SPY)",
    "Very High": "0% bonds (AGG), 100% equities (SPY)"
    }
    return risk[risk_level]

It follows that the initial_recommendation defined under the current recommend_portfolio(intent_request) function becomes:

# Get the initial investment recommendation
    initial_recommendation = recommended_portfolios(risk)

• Add additional slots to respond based on feedbacks

  • Allow customers to go back and re-select their risk levels
  • Provide more customized recommendations, e.g. more variety under each riskLevel

• Deploy RoboAdvisor on Facebook or Slack on Channels tab using Amazon Lex Documentations

  • Graphic analyzers are available under Monitoring tab on Lex

References

• Columbia University GitLab Repository
• https://docs.aws.amazon.com/lex/latest/dg/gs2-prepare.html
• https://docs.aws.amazon.com/lex/latest/dg/additional-exercises.html
• https://hackernoon.com/building-a-serverless-chatbot-with-aws-lex-lambda-and-amazon-aurora-part-1-5406e8db6123
• https://chatbotsmagazine.com/quick-start-develop-a-chat-bot-with-aws-lex-lambda-part-1-b6f7c80ebba6
• https://medium.com/velotio-perspectives/amazon-lex-aws-lambda-beyond-hello-world-1403c1825e72
• https://docs.aws.amazon.com/lex/latest/dg/example1.html
• https://docs.aws.amazon.com/lex/latest/dg/gs-cli-publish-bot.html
• https://docs.aws.amazon.com/lex/latest/dg/fb-bot-association.html
• https://www.iconfinder.com/
• CloudApp - Screen Recorder
• QuickTime Player