AI Policy For Credit Unions

Developed in partnership with leading credit unions

Download your free AI Policy now!

Download now!

You may unsubscribe from these communications at any time. By clicking download, you consent to allow Aviary to store and process the personal information submitted above to provide you the content requested.

Thanks for subscribing! A confirmation email will be sent to your inbox shortly. If you don't see it, please check your spam folder.
Oops! Something went wrong while submitting the form.

Edition 1: Another AI Revolution?

March 24 - 30: Will AI replace software engineers? What does Nvidia's biggest chips ever mean for financial institutions? Plus, a step by step guide on the easiest AI win for credit unions. All this and more included in the first ever edition of the AviaryAI newsletter.

Welcome to the first ever edition of the AviaryAI newsletter!

Thanks for joining us as we explore the intersection of GenAI and finance with practical learnings and the latest relevant insights. Let’s get started.

This week you’ll learn:

  • How good is the AI software engineer?
  • What Nvidia’s Blackwell chips could mean for your institution.
  • The easiest AI win for credit unions.
  • What McKinsey has to say about GenAI in Risk and Compliance

Subscribe to the AviaryAI Newsletter here.

The coolest things we're watching and why you should care
Graph displaying Devin's performance compared to other leading generative AI models

Will AI replace software engineers?
The AI software engineer, Devin correctly resolved 13.86% of issues with no human intervention. A groundbreaking result compared to previous models at 1.96%. - Read the full story here.
So what? Skilled engineers are still needed to understand the context of problems and potential solutions. But with Devin, every engineer you have is now worth 1.5 engineers. 

"AI-washing" is now a thing
The SEC recently fined two investment advisors for allegedly lying about their use of AI. - Read the full story here.
So what? With GenAI such a buzzword these days, confirming a partner’s AI capabilities is more important than ever. Past experience and results are great starting points.

Another AI revolution? 
Nvidia’s new line of “Blackwell” GPUs could make AI 2-5x faster.
- Read the full story here.
So what? AI runs on GPUs and Blackwell chips will result in smarter, faster, and more complex models than ever. As impressive GenAI capabilities are now, be on the lookout for even more value on the horizon.



Tokens are the basic building blocks of an LLM and allow it to process language. Consider the sentence: "The cat sat on the mat” LLMs need to break this sentence down into pieces it can process: "The","cat","sat","on","the","mat".These become the first set of tokens. Once language is “Tokenized” the LLM can analyze the relationships between tokens and begin to predict what tokens should come next given a certain context.


One simplified GenAI concept per week to build your AI Acumen

One Easy Quick AI Win For Credit Unions

Prepping your credit union’s knowledge base is a simple action that can dramatically increase your speed to AI adoption regardless of your eventual AI use-cases.

Your knowledge base is your institutions collection of internal and external documentation including:

  • FAQs
  • Procedures
  • Guidelines
  • Policies
  • Best practices.

With the rise of GenerativeAI, this knowledge base has never been more valuable. It’s the fuel that powers your AI. Like any engine, it needs proper maintenance and the right fuel to perform at its peak. AI models and techniques like retrieval augmented generation (RAG) process the information in your knowledge base to achieve the desired outputs.

Here are some steps to optimize your knowledge base for AI adoption:

Step 1: Declutter and Organize

✅ Toss outdated, dusty documents and policies.

✅  Categorize and label information for easy access.

✅  Weed out incorrect and inconsistent data. Accuracy enables AI success.

✅  Simplify jargon and flowery text. Your AI favors bullet points, clear headers, and concise language.

Step 2: Structure Content for AI

✅  Cluster closely related topics

✅  Identify key concepts and terms to guide AI to the right answers.

✅  Explain concepts clearly and concisely. AI thrives on focused data.

Step 3: Encourage Continuous Feedback

✅  Prompt members to rate answers, flag inaccuracies, and suggest improvements. This refines your data and sharpens your AI.

✅  Regularly review feedback and update your knowledge base to enable continuous AI learning

The latest news at the intersection of GenAI and Finance

Mckinsey’s Report - How GenAI can help banks manage risk and compliance

The report is chock-full of use cases with massive potential. We read it so you don’t have to:

McKinsey outlines three general use cases of generative AI

  1. A virtual expert a user can ask questions and receive answers sourced from long-form documents.
  2. Manual Process Automation where gen AI performs time-consuming tasks.
  3. Code acceleration where gen AI updates, translates old code or even writes new code.

Applying these 3 use cases to risk and compliance results in applications with a lot of potential.

  • In regulatory compliance: GenAI can automate regulatory compliance checks, monitor for potential breaches, and serve as a virtual expert to answer policy questions.
  • In financial crime: GenAI can generate suspicious activity reports, automate customer risk ratings, and improve transaction monitoring.
  • In credit risk: GenAI accelerates the end-to-end credit process by summarizing customer info, drafting credit memos, and generating risk reports.
  • In modeling and analytics: Legacy code migration, model performance monitoring, and documentation generation can be automated with GenAI.
  • In cyber risk: GenAI can check for vulnerabilities, generate secure code, and provide security insights.
  • In climate risk: assessment is streamlined as GenAI automates data collection, generates ESG reports, and assists risk visualization.

With so much potential, where do we start? McKinsey suggests prioritizing use cases with high impact, low risk, and high feasibility. The key considerations here include fairness, IP rights, privacy, security, explainability, strategic risks and 3rd party risks.

Read the full report here.

About AviaryAI

AviaryAI is the next evolution of financial interaction. Enhance your team with proactive outbound voice agents to welcome new members, encourage credit card activations, and drive non-interest revenue.

Drive Revenue and Engagement
With AI Today