Building next-gen GenAI solutions on Amazon Bedrock

Building next-gen GenAI solutions on Amazon Bedrock

Avannis, a leading provider of customer engagement solutions for banks and credit unions, found itself facing a significant challenge. With over 20,000 customer surveys to process each month, their team was struggling to keep up with the manual review, categorization, and classification of each survey.

The time-consuming nature of the task led to staff burnout and hindered the company's growth.


A challenge worth solving

Recognizing the need for change, Avannis partnered with our team at Innovative Solutions to explore the potential of Generative AI in transforming their operations. Innovative Solutions' Tailwinds architecture, built on Amazon Web Services (AWS) Bedrock, offered a promising path forward.


Tailwinds and Amazon Bedrock

Tailwinds is a technology solution bundle from Innovative that

  • Makes it easy for SMBs and startups to use generative AI on AWS.
  • It runs on Amazon Bedrock with pre-packaged AWS services and services from several AWS ISV partners such as Weaviate Anthropic IBM and Cohere

Leveraging the flexibility of Tailwinds and the power of Amazon Bedrock, Innovative Solutions worked closely with Avannis to develop a custom GenAI solution.

The solution, powered by Anthropic's Claude LLM via Amazon Bedrock, was trained on Avannis's extensive historical survey data, taxonomy requirements, and business rules.

Tailwinds greatly improved Avannis's survey processing workflow by generating semantically-similar surveys and requirements, enabling highly accurate classifications.

This architecture and process design took many iterations, going back and forth with Avannis to understand their accuracy requirements and acceptance of the output labels.

Ultimately, this was a novel approach to solving their problem. Alternatives could have included developing a custom ML model to label content.

Unfortunately, their data was not prepared to be leveraged for ML training and would have taken months to properly clean.

Instead, this solution was built, tested, and deployed in weeks (not months).

Meaningful impact starts here

The implementation of the Tailwinds GenAI solution yielded impressive results for Avannis.

  • Projected annual savings of $54,000 through reduced manual labor.
  • Staff morale improved significantly as employees were able to focus on more engaging and valuable tasks.


If you want to read the whole case study, check out the official Avannis case study page here.

Varsha K

Data Analyst at TalentMonk

9 个月

While Tallect [https://www.tallect.com/] isn't directly applicable to Avannis's challenge of processing customer surveys, their solution highlights the potential of AI for automating tasks and improving efficiency. Similar to how Avannis leveraged AI to streamline survey classification, Tallect uses AI to automate various HR processes, freeing up HR teams for more strategic work.

回复
Godwin Josh

Co-Founder of Altrosyn and DIrector at CDTECH | Inventor | Manufacturer

1 年

Starting a new series focused on GenAI implementations showcases a pivotal shift towards leveraging cutting-edge technology to streamline operations and enhance customer experiences. Considering the emphasis on scalability and work-life balance, how do you envision striking a balance between automation and human touch in customer engagement solutions like Avannis? Furthermore, could you share insights on the integration challenges and successes encountered while implementing GenAI on Amazon Bedrock, particularly in the context of white-glove customer service?

回复
Andrew Smith MBA

Chief Learning Officer @ Momentum Leadership | MBA

1 年

Excited to follow your journey with Built on Bedrock!

Nancy Ho

● Helping C-Level Execs, Mid-Level Managers, & Business Owners Bridge The Gap Between ???????????????????????? ?????????????? & ???????????????? ?????????????????????? ● Thought Leader on "The Professional Paradox"

1 年

Fascinating initiative! How does Avannis manage to improve work-life balance and growth?

回复
AJ Rivera

Managing Director - Head of Global Startups (2X Ex-Founder)??

1 年

Great work Travis Rehl ??

回复

要查看或添加评论,请登录

Travis Rehl的更多文章

  • Using AI to save $80,000/year with 2 hours of work

    Using AI to save $80,000/year with 2 hours of work

    In today's fast-paced business environment, efficiency is key. At Innovative Solutions, we recently faced a challenge…

    2 条评论
  • Turning GenAI POCs from Months to Minutes

    Turning GenAI POCs from Months to Minutes

    In the rapidly evolving landscape of Generative AI, technical leaders face a common challenge: “how do we quickly…

    5 条评论
  • Empowering Personalized Technical Training with GenAI: INE’s Success Story

    Empowering Personalized Technical Training with GenAI: INE’s Success Story

    Read the full case study here In the fast-paced world of IT, staying ahead requires continuous innovation. This is the…

    1 条评论
  • Unlock Data Insights with Semantic Labeling

    Unlock Data Insights with Semantic Labeling

    Many businesses today are looking for ways to analyze user/human generated content on their platforms. Are…

  • Building a Multi-Agent Orchestration Assistant

    Building a Multi-Agent Orchestration Assistant

    For those who are unfamiliar, Innovative Solutions has a long history of providing Managed Cloud Services to hundreds…

    6 条评论
  • Personalizing Content @ Edge with GenAI

    Personalizing Content @ Edge with GenAI

    Today's consumers expect relevant tailored content the moment they engage with a brand. As personalization becomes…

    1 条评论
  • Go beyond ETL with GenAI

    Go beyond ETL with GenAI

    ETL can be a pain in the %@#. Yet in the digital age, data is king.

  • What if: AWS DevOps powered by LLMs

    What if: AWS DevOps powered by LLMs

    Imagine you are driving home from work and you receive a Critical Alert! A system or environment is down and..

    2 条评论
  • Forget low-code. AI does it better.

    Forget low-code. AI does it better.

    In this article you will learn how Generative AI can replace or augment low-code Enable non-technical users to live out…

    4 条评论
  • Stop Babysitting ML Models

    Stop Babysitting ML Models

    Let's be honest. Labor is expensive, and business budgets are tightening.

    2 条评论

社区洞察

其他会员也浏览了