Adereso AI

What is Adereso AI?

Adereso AI is a full-channel conversational sales, customer service, and customer experience platform designed for medium to large enterprises. It integrates WhatsApp, social media, web chats, and email conversations into a single system, and utilizes AI agents in collaboration with human agents to handle pre-sales, transactional, and post-sales tasks.

The company’s headquarters are located in Chile; its products and documentation are primarily in Spanish. It serves teams in the e-commerce, retail, finance, logistics, and utilities sectors across Latin America. The company’s official website emphasizes charging based on the volume of conversations, rather than on the number of agents.

Main product components

Main functions

• Centralized management of WhatsApp, social media, web chats, and email.
• Use AI agents to answer pre-sales and after-sales questions.
• Recommend products during conversations, address objections, and encourage purchases.
• According to the rules, complex conversations are routed to the appropriate human agent.
• Maintain the complete context of the conversation between the robot and the human.
• Create tickets, queues, departments, assignments, and service processes.
• Send WhatsApp marketing and transaction notifications.
• Restore abandoned purchases, reactivate leads, and encourage repeat purchases.
• Analyze NPS, CSAT, customer sentiment, and service trends.
• Connect to Shopify, Salesforce, HubSpot, and enterprise systems.

Omni-channel help desk

Adereso Desk converts messages from various channels into unified sessions and tickets. Agents do not need to switch between different applications, and tasks can be assigned based on the customer, channel, topic, priority, and team.

• The unified inbox displays messages from customers across various channels.
• Organize pending tasks by column, queue, and department.
• Set automatic allocation, priorities, and service rules.
• Use preset replies and AI suggestions to increase processing speed.
• Retain the historical context when the customer contacts again.
• View the agent’s workload, response time, and resolution status.
• Deleting a work view does not automatically remove the underlying ticket details.
• It supports an unlimited number of seats, but the total cost is still influenced by the volume of conversations.

AI sales agent

Sales bots can not only answer FAQ, but also recommend products based on customer needs, create shopping carts, provide payment links, and handle common objections. They are suitable for brands with clear product catalogs, well-developed online payment systems, and a high volume of WhatsApp traffic.

• Understand the customer’s needs and ask about the necessary filtering criteria.
• Recommend products based on catalog, price, and inventory.
• Explain the specifications, delivery, return and exchange, as well as promotional rules.
• Create a shopping cart or guide customers to the checkout page.
• Perform cross-selling, upgrade selling, and reorder reminders.
• Identify high-intent leads and pass them on to sales staff.
• Save the source, intent, products, and session summary.
• Review prices, inventory, and commitment usage with real-time data.

AI after-sales agent

• Answer questions regarding order status, delivery, and pickup.
• Explain the return, exchange, warranty, and service policies.
• Query customer transaction details based on the order number.
• Handles business hours, store locations, and common operation instructions.
• Create services or refund requests within the scope of permissions.
• Identify complaints, fraud, urgent cases, and high-value customers.
• If it cannot be resolved, please submit it along with a summary and context to human support.
• Manual approval and audit records are retained for sensitive decisions.

Adereso Studio

Adereso Studio is a no-code tool for dialog automation; teams can configure AI robots, personalities, knowledge bases, and business processes. The help center also provides instructions on testing, accounts, versions, processes, and integration with Shopify robots.

• Create robots with clear roles and a brand voice.
• Import FAQs, policies, product information, and business knowledge.
• Design issues, conditions, actions, and manual workflow transitions.
• Use a tester to verify different inputs before going live.
• Manage robot versions and save change logs.
• Connect the robot to an event or service channel.
• Use generative AI to assist in creating and optimizing agents.
• Provide safety exits for knowledge gaps and uncertain answers.

Human collaboration with AI

Adereso views AI as a tool for automating repetitive tasks and enhancing the work of customer service agents, rather than as a means to completely replace customer service teams. Robots can transfer a request to another department or specific employee within the same conversation, so customers do not need to restate their issues.

• AI first handles business hours, logistics, policies, and basic queries.
• Business rules determine when a case must be transferred to a human agent.
• Historical messages, customer details, and summaries are included when transferring.
• Humans can continue the original session and modify the AI suggestions.
• Once resolved, automation can continue with notifications and satisfaction surveys.
• Managers can review failed sessions to improve knowledge and processes.
• Key responsibilities for sales, complaints, and exception handling remain with the assigned personnel.

WhatsApp marketing and proactive outreach

• Create message templates that comply with WhatsApp’s rules.
• Build audiences based on customer attributes, purchase history, and behavior.
• Send transaction notifications such as order placement, delivery, and reservations.
• Run promotions, restocking, repeat purchase, and recall campaigns.
• Drive traffic from ads, websites, and social media to WhatsApp.
• Carry out controlled follow-ups for leads that have not been purchased.
• Track performance in terms of sends, replies, conversions, and unsubscribes.
• Send marketing messages only to users who have given appropriate consent.

Restoring abandoned purchases and encouraging repeat purchases

Solutions at the Scale level and above focus on automatic recovery of abandoned purchases; the system can initiate personalized conversations after the customer leaves the checkout process. Businesses can also carry out follow-up sales and repeat purchases by taking into account the replenishment cycle, related products, and the customer’s past interests.

• Identify shopping carts and customers with unfinished settlements.
• Set trigger conditions based on time, value, and product.
• Send reminders with a shopping cart or checkout link.
• Answer common questions that cause customers to give up on their purchases.
• Pass high-value opportunities to sales staff when necessary.
• Limit the frequency and respect cancellation and silence periods.
• Evaluate performance using incremental conversion rather than simple clicks.

Clue filtering and reservation

• It automatically asks about the budget, requirements, location, and purchase time.
• Distinguish between information inquiries, low-intent leads, and qualified leads.
• It is allocated to the sales team based on scores and rules.
• Synchronize contacts, sources, and conversation summaries to CRM.
• Check the calendar and schedule meetings for qualified clients.
• Provide other resources to users who do not meet the requirements.
• Discriminatory screening based on sensitive attributes is prohibited.

Customer feedback and VoC analysis

Adereso Experience is used to gather customer feedback, combining structured ratings with textual comments for analysis. Enterprise also includes VoC Analytics, which helps management identify issues, emotions, and trends across different channels.

• Automatically send NPS and CSAT surveys.
• Analyze the results by channel, product, store, and team.
• Use AI to identify positive, negative, and neutral emotions.
• Clustered repeated complaints, as well as frictions related to demands and services.
• Abnormal trends were detected in the real-time dashboard.
• Generate reports for management and business units.
• Route serious feedback to the responsible personnel.
• Emotional inference may misjudge irony, dialects, and context.

Support channels

CRM integration with e-commerce

The official website specifies the native integrations with Salesforce, HubSpot, and Shopify, and states that the technical team can develop custom connections for other systems. The scope of integration and the direction of synchronization must be confirmed in writing during the procurement and implementation phases.

• Read product, order, and cart information from Shopify.
• Write contact information, intentions, activities, and session summaries to CRM.
• Use customer attributes for routing and personalization.
• Synchronize the results of manual services back to the business system.
• Connect to the Google Drive knowledge base.
• Integration with other systems is possible through APIs or custom development.
• Set minimum permissions and error retry options for each connection.

Appropriate use cases

• E-commerce businesses use WhatsApp to recommend products and complete sales.
• Retail brands consolidate pre-sales and after-sales services across multiple channels.
• Logistics companies automatically answer questions regarding package tracking.
• Public utilities handle billing, services, and inquiries regarding issues.
• Financial and insurance companies identify leads and schedule consultations with advisors.
• Large customer service centers automatically handle frequently recurring issues.
• The brand carries out activities related to abandoned purchases, repeat purchases, and customer recall.
• The CX team analyzes NPS, CSAT, sentiment, and complaint trends.

Which companies are suitable?

• WhatsApp is the primary channel for communicating with customers in Latin American companies.
• Teams that have thousands of service or sales conversations each month.
• Organizations that require unlimited seats and the ability to scale based on the volume of conversations.
• Companies that hope for AI to work in collaboration with humans rather than having robots operate in isolation.
• Large and medium-sized brands that possess product catalogs, CRM data, or e-commerce data.
• Companies that require Spanish-language implementation and ongoing optimization support.
• A team capable of managing automation, privacy, and marketing consent.

Which scenarios are not suitable?

• Only a small number of individuals and micro-enterprises seek consultation each month.
• Teams that only need a simple web FAQ plugin.
• Companies in key markets that do not use Spanish and require full support in Chinese.
• Users who require complete self-service and do not need any service provision.
• A sales robot project that is unable to connect to customer, product, and order data.
• Organizations that seek a fully open-source and self-hosted platform as a whole.
• Marketing teams that are unwilling to comply with WhatsApp’s templates, terms of service, and messaging policies.

Adereso Implementation Tutorial

1. Choose a measurable goal, such as response time, automatic resolution, or WhatsApp sales.
2. Audit channels, volume of conversations, agents, CRM, e-commerce, and sources of knowledge.
3. Define the boundaries for what AI can handle, what needs to be confirmed, and what must be referred to human staff.
4. Connect the WhatsApp Business account to priority publishing channels.
5. Import knowledge, product catalogs, policies, and representative historical issues.
6. Configure the robot’s tone, processes, departments, and assignment rules.
7. Run normal, abnormal, and high-risk sessions in the testing environment.
8. After training the agents, the system is launched on a limited scale, with continuous review of failed samples.

WhatsApp sales process tutorial

1. Identify the legitimate sources of traffic for advertisements, websites, or customer lists.
2. Different entries are designed for inquiries, recommendations, the shopping cart, and after-sales support.
3. Connects real-time product, price, inventory, and settlement information.
4. Let AI first understand the requirements, and then recommend a small number of relevant products.
5. Confirm the key parameters before sending the shopping cart or payment link.
6. Set up manual routing for price exceptions, complaints, and large orders.
7. Configure purchase abandonment alerts, frequency control, cancellation options, and silent periods.
8. Incremental orders, revenue, and customer satisfaction are measured against a control group.

Tutorial on robot quality inspection

1. Collect real FAQs, order-related questions, as well as spoken and spelling variations.
2. Create a list of correct answers, permitted sources, and prohibited commitments.
3. Test whether uncertainty is acknowledged when knowledge is lacking, and switch to a human operator.
4. Verify whether the context remains correct after the customer switches channels.
5. The test alerts for injection, abuse, fraud, and the entry of sensitive information.
6. Check whether the model makes incorrect recommendations, invents inventory levels, or promises refunds.
7. Perform regression testing for each update to knowledge or processes.
8. After going live, random checks are conducted on sessions with low ratings, those transferred to human agents, and complaint-related sessions.

Price packages

The official website currently offers three plans: Growth, Scale, and Enterprise, with the option to switch between annual and quarterly billing; it is stated on the page that the annual plan can save 40% in costs. The table below shows the starting prices in dollars as displayed on the page at the time of verification; taxes, WhatsApp fees, and custom integrations may incur additional costs.

Comparison of package features

Excess and activity costs

It is necessary to confirm before making a purchase.

• Is the quote based on an annual or quarterly commitment, and what is the corresponding actual average monthly cost?
• Statistical definitions for centralized sessions and automated sessions.
• Are the costs for WhatsApp and other channels included?
• Robot invocation of models, knowledge bases, and message volume limits.
• There are separate costs and maintenance responsibilities associated with native integration and custom development.
• How many man-hours are involved in implementation, training, testing, and continuous optimization?
• Data retention, export, deletion, and migration after the contract ends.
• Service level, fault response, and regional data processing requirements.

Safety and privacy

• Only the minimum amount of customer data necessary for sales and service is collected.
• Use the minimum required permissions for CRM, e-commerce, and knowledge base connections.
• Payments, refunds, account changes, and sensitive information require enhanced verification.
• Marketing campaigns must document consent, unsubscrictions, and template approvals.
• Review the model providers, as well as the mechanisms for logging and data retention.
• Hide ID numbers, card numbers, and unnecessary personal information.
• Restrict agents from accessing customers and tickets by department.
• Maintain audit records of changes to robots, human intervention, rules, and data.

Effect evaluation indicators

• First response time, average response time, and resolution time.
• AI automatic resolution rate and correct handover to human agents rate.
• The conversion rate from sales conversations to carts and orders.
• The additional revenue generated by the resumption of discontinued purchases.
• Each conversation reduces costs and saves time.
• CSAT, NPS, complaints, and repeat contact rate.
• Rates of delivery, response, cancellation, and blocking for messages.
• False promises by AI, privacy breaches, and high-risk mistakes.

Product advantages

• Sales, customer service, events, and customer experience form a complete closed loop.
• In-depth understanding of WhatsApp and its application in Latin American business environments.
• Omni-channel sessions and human agents are centralized on the same platform.
• When switching from AI to human assistance, the context is retained, reducing the need for customers to repeat their explanations.
• Unlimited seats are suitable for large-scale customer service and cross-departmental collaboration.
• It supports core systems such as Shopify, HubSpot, and Salesforce.
• It provides implementation, training, and ongoing business review.
• The public prices and the unit cost per additional session facilitate preliminary budgeting.

Usage restrictions and precautions

• The minimum monthly fee and implementation cost are more suitable for medium to large enterprises.
• Products, documents, and services are primarily aimed at the Spanish-speaking market.
• The total cost may also include fees for the WhatsApp channel and custom integrations.
• When billing is based on sessions, activities and peak traffic can quickly increase costs.
• AI may fabricate prices, inventory levels, policies, or processing results.
• Multi-channel integration increases the complexity of permissions management, synchronization, and troubleshooting.
• Sales automation must comply with rules regarding consent, cancellation of subscriptions, and consumer protection.
• The examples on the official website are supplier-specific data, and they need to be verified through one’s own experimental tests.
• Critical operations still require manual supervision and clear fallback procedures in case of failures.

GitHub and the open-source status

Adereso has a public GitHub organization, where it has released Python-based implementations of the Desk interface, various interface components, and example projects. The full versions of Desk, Studio, Engage, Experience, and the AI platform remain closed-source commercial services.

Basic information

Recommendation score

4.5 / 5. Adereso AI integrates WhatsApp sales, omnichannel services, the collaboration between AI and human agents, event management, and VoC analysis to form a comprehensive system that is suitable for medium to large enterprises in Latin America; however, the implementation costs and monthly fees are high, and the overall costs and operational risks increase as the volume of conversations, the number of channels, and the complexity of integrations rise.

Frequently Asked Questions

What is Adereso AI mainly used for?

It aggregates customers’ digital sessions and utilizes both AI and human expertise to carry out product recommendations, sales, after-sales support, marketing activities, and customer experience analysis.

Which channels are supported?

It supports WhatsApp Business, Instagram, Facebook Messenger, X, LinkedIn, web chat, and email.

Can it be connected to Shopify?

Yes, both the official website and the help center provide information on Shopify integration; integration with Salesforce, HubSpot, and custom connections is also supported.

What to do when AI cannot answer?

The system can transfer a customer to a designated department or agent within the same session, while passing on the complete history and context as well.

Is it charged per agent?

The official website states that fees are calculated based on the volume of conversations rather than the number of users, and all packages allow for an unlimited number of agents.

How much is Adereso AI?

The implementation fee and monthly fee for Growth start at $558; the implementation fee for Scale is $1,600 with a monthly fee starting at $1,058; Enterprise offers customized options.

How many conversations are included?

Growth includes 2000 instances, while Scale includes 5000 instances; excessive concentration sessions and automated sessions are charged at different rates.

How long will it take to implement?

Officials usually provide a timeframe of 2 to 4 weeks, which includes objectives, integration, training, testing, as well as ongoing optimization after the system goes live.

Is Adereso an official partner of Meta?

The official website refers to it as a WhatsApp Business Solution Provider, one that can assist companies with account verification, the use of templates, and integration with corporate systems.

Is Adereso AI open source?

The entire platform is not open-source, but the official GitHub page provides Python interface wrappers, UI components, and a few example projects.