Genesys Cloud - Main

Hello, @Victor Soares

In my experience, the main AI capabilities I'm currently working with in Genesys are Agent Copilot and bots.

For Agent Copilot, one of the biggest challenges is not just enabling the feature, but first building a solid foundation behind it. A good Knowledge Base is essential, with well-structured articles, clear content, and good intent mapping. If the knowledge and intents are not properly designed, Copilot may not provide the best recommendations to agents. So, a big part of the work is organizing, tuning, and continuously improving the Knowledge Base and intents to make the assistance more accurate and useful.

For bots, I see a lot of value in journey design. It is important to understand the customer's real path and decide when the bot should answer with FAQ functionality, when it should guide the customer through a flow, and when it should transfer to an agent. Implementing FAQ features can help reduce simple and repetitive contacts, but the experience needs to be well designed so the customer does not feel stuck in the bot.

Going forward, I would like to see more native reports for Agent Copilot. Today, reporting is very important to show customers the real value of the solution, identify improvement opportunities, and make their lives easier when analyzing results. Having more built-in reports would help teams track performance, understand usage, and continuously improve the Copilot experience without depending so much on custom reports.

We are currently exploring and using AI capabilities in Genesys mainly to improve customer self-service, agent productivity, and operational visibility.

Some of the most valuable use cases include:

• Bots and digital flows to automate common requests and reduce repetitive interactions for agents.
• Intent recognition and routing to better understand customer needs and direct conversations to the right queue or specialist.
• Agent assist / knowledge suggestions to help agents find the right information faster during interactions.
• Conversation summaries to reduce after-call work and improve handoff quality.
• Sentiment and interaction analytics to identify pain points, recurring issues, and opportunities to improve the customer journey.

These capabilities have helped solve challenges such as long handling times, inconsistent responses, high transfer rates, and the difficulty of analyzing large volumes of interactions manually.

Going forward, I would like to see even stronger AI features in Genesys around:

• More flexible and accurate generative AI for bots
• Better native integration between AI, Architect flows, scripts, and Data Actions
• Improved multilingual support, especially for Portuguese
• Easier tools to train, test, and monitor AI models
• More advanced real-time agent guidance
• Better AI-driven insights for identifying automation opportunities

Overall, I see AI in Genesys as a way to make customer interactions faster, more personalized, and more consistent, while also giving agents better tools to focus on complex cases.

Hi Victor,

We are currently using several AI capabilities in Genesys Cloud, mainly around Agent Copilot, knowledge surfacing, summarisation, sentiment/topic analysis, and automation workflows.

Some of the biggest benefits we've seen are:

• faster agent onboarding and guidance through surfaced knowledge/articles
• reduced manual effort with AI-generated conversation summaries
• better visibility into customer sentiment and emerging issues
• identifying coaching/training opportunities through topic detection
• improved self-service experiences through bot and knowledge integrations

We've also implemented workflow-based notifications using Speech & Text Analytics topic and sentiment triggers combined with AI Insights summaries. This has been particularly useful for surfacing potential complaints, dissatisfaction trends, and possible knowledge gaps much faster to operational teams.

We will also soon be looking into AI-powered auto-evaluation use cases within QA to help further streamline quality and coaching processes.

One area I'd personally love to see continue improving is deeper cross-platform integration and orchestration, especially around:

• AI-driven workflow actions
• proactive recommendations
• richer Copilot APIs
• more configurable summarisation outputs
• tighter CRM/ITSM integration patterns
• broader automation capabilities from conversation insights

Would also love to see even more flexibility around AI-powered quality management and auto-evaluation scenarios in the future.

We are currently exploring and using AI capabilities in Genesys mainly to improve customer self-service, agent productivity, and operational visibility.

Some of the most valuable use cases include:

• Bots and digital flows to automate common requests and reduce repetitive interactions for agents.
• Intent recognition and routing to better understand customer needs and direct conversations to the right queue or specialist.

Going forward, I would like to see even stronger AI features in Genesys around:

• More flexible and accurate generative AI for bots
• Better native integration between AI, Architect flows, scripts, and Data Actions
• Improved multilingual support, especially for Portuguese

Hello @Victor Soares!!

In my day-to-day work with Genesys Cloud, the AI feature that has made the biggest difference for me is Agent Copilot. It helps agents stay focused on the conversation instead of searching through multiple systems, and the automatic summaries save a lot of time after each interaction.

I've also seen great results with Predictive Routing and Virtual Agents/Bots. Predictive Routing helps connect customers with the best available agent, while bots handle repetitive requests and allow agents to focus on more complex situations. Both have contributed to improving customer experience and operational efficiency.

• More powerful and context-aware Agent Copilot capabilities.
• Better support for multilingual environments.
• AI that can proactively recommend operational improvements, not only for agents but also for supervisors and workforce management teams.
• Deeper integration between AI insights and business applications, enabling more automated actions directly from the interaction context.

Overall, AI in Genesys is already helping us work smarter rather than harder, and I'm excited to see how the platform continues to evolve over the next few years!!!

------------------------------
Vinicius Campos
Solutions Architect
------------------------------

Hi @Victor Soares,

We are currently exploring and using AI capabilities in Genesys mainly to improve customer self-service, agent productivity, and operational visibility. Regarding the Copilot ecosystem:

Current Impact:

• Agent Copilot: The primary gain has been the reduction in manual effort for knowledge retrieval and real-time summarization. By automating the extraction of context from previous interactions, agents are able to handle complex queries with significantly lower cognitive load, directly improving handle time and accuracy.

• Supervisor Copilot: The shift from reactive monitoring to real-time, AI-driven guidance has been crucial. It allows for faster identification of performance anomalies during live calls, enabling supervisors to provide timely support without having to manually sift through massive amounts of interaction data.

• Admin Copilot: We have been utilizing it to streamline routine administration tasks. This has been particularly effective in reducing the time required to configure and update routing rules, which helps maintain system stability while allowing for faster deployment of configuration changes.

Future Expectations & Improvements:

• Prescriptive Administration: For the Admin Copilot, the next step is moving toward a 'pre-flight' validation model. It would be highly valuable to have an AI that runs impact analyses on routing or workflow changes before they go live, predicting potential performance drops or configuration conflicts based on historical data.

• Actionable Supervisor Insights: I would like to see the Supervisor Copilot evolve beyond identifying issues. It should be capable of suggesting-or even autonomously executing-real-time adjustments to workforce distribution or queue settings when specific performance thresholds are violated, moving the platform toward true self-optimizing operations.

------------------------------
Leonardo Vieira
Telecom Specialist
------------------------------

We are currently exploring Agentic Virtual Agent capabilities, and I see a lot of potential there.

What excites me most is the possibility of going beyond traditional bots that only answer questions, and moving toward AI that can take actions, follow business rules, call APIs, and help solve customer needs more end-to-end.

Since this is still a newer capability, there is a lot to test and validate, but that is exactly what makes it so interesting.

Going forward, I would like to see more evolution around orchestration, testing, visibility into AI decisions/actions, and better connection with business rules and reporting.

I have been using AI Powered Slots mainly to handle more complex data collection scenarios within bot flows, such as alphanumeric codes.

Key benefits I have experienced:

• The process of capturing and processing this type of information became much simpler and more efficient.

• It significantly reduced the complexity of validations within the bot flow.

• It improved the user experience during interactions with the bot.

• It made the overall bot flow development easier and brought more accuracy to data collection.

As a point for improvement:

• I believe the current limit for example utterances is still somewhat restrictive.

• Having the ability to add more examples would greatly help with model training.

• It would also improve recognition accuracy in more varied and complex scenarios.

Hello,

We are currently exploring the Agentic Virtual Agent. This is very exciting because we started with the traditional bot that has some limitations, but the AVA will make it feel more natural.

One of the challenges though is knowing the exact structure that the VA consumes the Knowledge base articles and the guardrails to prevent it from getting outside information. Also, even without a knowledge base connected it tries to answer questions.

Hi all,

thank you for sharing!

I'm currently collecting experiences with Agentic Virtual Agent and was searching for a guide how to prompt the best way for the AVA. Do you have any resources helping here out? This is something which I miss for AVA.

@Valentina Postelnicu @Georgios Kondylas #AdvisoryServices

A great topic to follow - such interesting stories