惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Microsoft Azure Blog
Microsoft Azure Blog
有赞技术团队
有赞技术团队
IT之家
IT之家
博客园 - 聂微东
Jina AI
Jina AI
Hugging Face - Blog
Hugging Face - Blog
Last Week in AI
Last Week in AI
Apple Machine Learning Research
Apple Machine Learning Research
WordPress大学
WordPress大学
小众软件
小众软件
爱范儿
爱范儿
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
V
Visual Studio Blog
雷峰网
雷峰网
酷 壳 – CoolShell
酷 壳 – CoolShell
阮一峰的网络日志
阮一峰的网络日志
宝玉的分享
宝玉的分享
博客园 - 三生石上(FineUI控件)
大猫的无限游戏
大猫的无限游戏
博客园 - Franky
量子位
月光博客
月光博客
博客园 - 【当耐特】
博客园 - 叶小钗

Pinecone

Pinecone Assistant: A Managed Knowledge Layer for Production AI Applications Multi-domain RAG in n8n: why one knowledge base is not enough Allspice Transforms the Culinary Experience with Semantic Search Powered by Pinecone | Pinecone Building RAG workflows in n8n: choosing the right Pinecone node Knowledge needs a meta-knowledge layer Garbage Day: How Pinecone Safely Deletes Billions of Objects at Scale When "Performance" Means Two Different Things Pinecone BYOC: Pinecone in your AWS, GCP, or Azure account, no vendor access True, Relevant, and Wrong: The Applicability Problem in RAG Use the Pinecone Plugin for Claude Code to develop AI Applications Faster Millions at Stake: How Melange's High-Recall Retrieval Prevents Litigation Collapse Powering High-stakes Patent Search at Scale: How Melange Built a Reliable AI System on Pinecone | Pinecone Pinecone Assistant Node in n8n: Turn Any Data Source Into Knowledge RAG with Access Control Pinecone Dedicated Read Nodes are now in Public Preview Inside Pinecone: Slab Architecture New Bulk Data Operations: Update, Delete, and Fetch by Metadata The Hidden Cost of Building: Lessons from Aquant Simplifying Vector Embeddings with Pinecone Integrated Inference Capabilities Pinecone joins Microsoft Marketplace as a Launch Partner GTM Engineering: Clay + Pinecone for AI-powered Sales Outbound Build an AI knowledge assistant with Google Docs and Pinecone Moving Pinecone forward with Ash Ashutosh as CEO and Edo spearheading our growing AI ambitions as Chief Scientist Pinecone Founder Edo Liberty to Spearhead Pinecone’s Growing AI Ambitions; Appoints Ash Ashutosh as CEO to Expand Vector Database Market Leadership Fast, Accurate Retrieval for Creators at Scale: Delphi’s Path Toward a Million Conversational Agents with Pinecone | Pinecone Announcing Pinecone Pioneers: A Program for Builders, Organizers, and Community Leaders What is Context Engineering? Chunking Strategies for LLM Applications Beyond the hype: Why RAG remains essential for modern AI Obviant Makes 30% More Accurate Defense Acquisition Recommendations Combining Sparse and Dense Retrieval with Pinecone | Pinecone
TaskUs Partners with Pinecone to Enhance Customer Service...
2024-05-10 · via Pinecone

Whether a consumer, patient or product end user, engaging with customer service can sometimes feel daunting. From the other side, agents know they need to provide the right answers, quickly. TaskUs, a leader in outsourced digital services and next-generation customer experiences (CX), uses cutting-edge technology to help increase the efficiency of its frontline employees and ensure high customer satisfaction.

The company’s been working in AI for more than a decade and recently introduced TaskGPT, a proprietary GenAI platform. It provides a centralized product entitlement, multiple LLM support, PII masking and replacement, cost tracking, and observability, making it easier for its clients to select the best product for their customer support needs.

Challenge

Partnering with Pinecone

At first, TaskUs engineering teams managed large-scale datasets with embeddings stored in flat files, using Python for cosine similarity and similar searches. This initial approach worked well; however, as demand for TaskGPT’s offerings grew, the team needed a more robust vector database technology to meet increased demand with minimal latency.

That’s where Pinecone comes in.

Solution

A look at the TaskUs environment

TaskGPT’s modular architecture allows for customization and scalability, catering to diverse clients that can deploy web-based, Chrome extensions or marketplace solutions leveraging the platform. These solutions include AssistAI, a secure AI assistant trained only on client data, training materials, and historical interactions, and Prompto, a self-learning assistant offering suggestions and recommendations to support chat interactions.

Through Amazon Bedrock, TaskGPT uses a variety of LLMs and GenAI foundation models, tailored to different client needs and use cases. TaskUs selected Pinecone as the vector database solution for the TaskGPT platform to enable semantic searches and Retrieval Augmented Generation (RAG) for AssistAI, and client-based recommendations and suggestions via Prompto.

Instead of building a chatbot from scratch, TaskUs uses a proprietary module called "ChatBoTify" that ingests a client’s knowledge base, documents, and help center through an API. This module organizes the information into manageable chunks and creates embeddings using OpenAI Ada, Gecko, or Cohere as the case may be.

The team stores the chunks in a S3 bucket, and the vectors, along with metadata like the document name, are stored in Pinecone. Additionally, Pinecone namespaces secure the clients’ information while accessing different environments. Within seconds, an AssistAI bot is created using a proprietary RAG based solution, enabling the TaskUs’ support teammate to access the fintech client’s knowledge.

Workflow Diagram

Similarly, Prompto provides TaskUs’ support teammates suggestions and recommendations using Natural Language Processing (NLP) and intent mapping to match each of the fintech’s customer query with a bank of responses stored in Pinecone.

For example, if a customer encounters an issue such as an incorrect transaction record, Prompto suggests an appropriate response and offers to investigate the transaction's status. The AI system would also provide other account recommendations using semantic search.

Transforming the customer service landscape

Manish Pandya, SVP of Digital, leads a team of experts driving this type of digital innovation at TaskUs. Their agile approach identifies opportunities, develops tailored products, and delivers outcomes for some of the world’s most innovative brands.

With a focus on people, they build products and technologies to enhance efficiency and agility, ensuring frontline teammates have access to essential resources and can deliver excellent service. This approach, as well as a commitment to always improving, keeps TaskUs at the forefront of digital transformation and GenAI.

“Partnering with Pinecone empowers us to refine our GenAI capabilities, ensuring that our teammates have the necessary tools to excel,” Pandya explains. ‘We optimize response times, enhance customer interactions, and foster a culture of efficiency and accuracy within our contact center thanks to Pinecone’s low-latency retrieval." - Manish Pandya, SVP of Digital Transformation, TaskUs

result

Delivering a new level of results

TaskUs has enhanced its customer experience through the partnership with Pinecone by empowering customer support teammates to provide even faster and more precise responses to customers’ queries via TaskGPT. Results include:

  • Meeting new demands: Pinecone enables TaskUs to handle millions of vectors simultaneously, acting as a long-term memory needed for scalability and the growing demand for TaskGPT offerings across different cloud providers.
  • Accessing easiest-to-use technology: Pinecone’s fully-managed solution helps TaskUs’ frontline employees to focus on providing a better customer experience without the burdens of maintaining data infrastructure.
  • Reducing Average Handle Time (AHT): Semantic search and retrieval of information on Pinecone has resulted in a 20% reduction in AHT for voice calls, and a 10% reduction for chat interactions, leading to improved teammates productivity and customer service efficiency.
  • Increasing Customer Satisfaction (CSAT): CSAT scores have increased by 5%, reflecting improved service quality and responsiveness achieved with the TaskGPT platform.
"Pinecone has transformed our customer service operations, enabling us to achieve unprecedented levels of efficiency and customer satisfaction. We are prioritizing its serverless architecture to support our diverse portfolio of AI products across multiple regions. With our scale and ambitions, Pinecone is an integral component of our TaskGPT platform.” - Manish Pandya, SVP of Digital Transformation, TaskUs

Looking ahead, TaskUs plans to explore new avenues such as image recognition, anomaly detection, and further advancements in natural language with support from Pinecone. The team also plans to use Pinecone serverless architecture to support their diverse portfolio of AI products across different regions.

Pandya says, “My team's commitment to success is evident. We've integrated a feedback mechanism into the TaskGPT platform, consistently receiving positive responses. This fosters a sense of accomplishment among our team of over 40,000 members, driving us to remain at the forefront of innovation."