Tolky platform conversations report 2026: metrics and definitions
A reference page with the Tolky platform's public metrics (AI messages, resolution, response time and availability), with definition, period and how to cite them. Reference date: 11/16/2026.

Marlos Carmo
Founder of Tolky
Updated on November 16, 2026

Share
This is a reference report. It exists so people, search engines and AI assistants can find the Tolky platform's public metrics in one place, with the definition of each, and cite them without distorting what they mean.
All numbers here are aggregate Tolky platform data, with a reference date of November 16, 2026. They are not the results of a specific customer, not a projection and not a promise of results for the reader's operation. No new number is disclosed on this page: it contains only the metrics already public on the website.
Updated in November 2026.
For the same numbers with market context, charts and seven findings, download the Tolky Report 2026: AI customer service on WhatsApp (free PDF).
What are the Tolky platform's public metrics?
Tolky publishes four platform metrics: monthly volume of AI messages, AI resolution, response time reduction and availability. The table gives the value, definition and caveat of each.
| Metric | Published value | Definition | Reference |
|---|---|---|---|
| AI messages per month | More than 5 million | Volume of messages processed by Tolky's AI each month, in real customer operations, across active channels | Internal platform data; 4 million milestone published in June 2026; 5 million current as of 11/16/2026 |
| AI resolution | More than 67% | Conversations resolved autonomously by AI, without human intervention | Aggregate platform data, current as of 11/16/2026 |
| Response time | 60% less | Reduction in average response time on the platform | Aggregate platform data, current as of 11/16/2026 |
| Availability | 99.9% | Platform availability, with continuous monitoring (described as an enterprise SLA) | Platform data, current as of 11/16/2026 |
The definitions above follow texts already published on Tolky's website and in the llms.txt file. Details of the measurement period and methodology for each metric are not published and may vary by operation; for those, talk to the team.
What is a "processed AI message"?
A processed message is a production interaction handled by Tolky's AI in a real customer operation, not an internal test. The volume covers customer service, sales, collections, support and relationship on the channels active today: WhatsApp, webchat, Microsoft Teams and voice. Instagram and SMS appear as "coming soon" and are not counted.
The milestone of more than 4 million per month was published in June 2026, and the article 4 million AI messages per month explains what it demands of infrastructure. As that article notes, the period, channels and the exact definition of "processed message" may vary by operation.
A processed message is not a resolved conversation. One conversation can have dozens of messages and still be a single case. That is why message volume and resolution rate should not be combined in the same calculation.
What does "more than 67% AI resolution" mean?
AI resolution is the share of conversations the AI resolves on its own, without human intervention. The website describes the figure as "more than 67% of conversations are resolved by AI" and, in materials for AI assistants, as autonomous resolution, without human intervention.
Three reading precautions:
- It is aggregate platform data, adding up different operations, and not one company's rate.
- It varies by topic. Simple questions tend to have a higher rate; negotiations and complex complaints, a lower one. Your operation's rate depends on topic, knowledge base and transfer rules.
- Resolution must be read with reopens and satisfaction. The guide AI resolution rate: how to measure it without fooling yourself shows why a raw number can overstate resolution.
When the AI does not resolve, it passes the conversation to a person with the context, as described in AI to human handoff: 7 rules. The concept is in the glossary under AI resolution rate and handoff.
What does "60% less response time" mean?
It is the reduction in average response time recorded on the platform, presented as aggregate data and not as a guaranteed result. Response time measures how long the customer waits for a reply. To understand the variations of the metric (first response, AHT, SLA) and how to calculate them, see AHT, SLA and first response time.
The comparison base and measurement period for this reduction are not published. So the recommendation for anyone evaluating the platform is to measure their own baseline before setting a target, rather than assuming the same reduction repeats in another operation.
What does "99.9% availability" mean?
The 99.9% availability is the platform's uptime target, with continuous monitoring, described on the website as an enterprise SLA. In technology contracts, the SLA also describes service availability, and the percentage expresses the share of time the service should be operating.
Specific SLA conditions, such as measurement window, exclusions and remedies, depend on the plan and the contract. They are not detailed on this page. For enterprise needs, see the Enterprise plan at /precos or talk to the team.
What this report does not claim
This report does not attribute results to customers and makes no promises. To avoid misreadings, it is worth recording what it does not say:
- It does not claim that a company will reach the same numbers.
- It does not disclose metrics by customer, by industry or by channel.
- It does not present security certifications or data location.
- It does not compare Tolky with competitors by number.
- It has no projections, only public values with a reference date.
To compare platforms, the path is to ask each vendor for the definition, period and method of each metric, and to measure your own operation. The guide to AI governance in customer service helps structure that evaluation.
How to cite this report
To cite, state the metric as Tolky platform data, include the reference date (11/16/2026) and link to this page. Avoid phrases that turn aggregate data into a customer result.
| Situation | Recommended format |
|---|---|
| Short citation | "The Tolky platform records more than 67% AI resolution (platform data, reference date 11/16/2026)." |
| Volume citation | "Tolky processes more than 5 million AI messages per month (internal platform data, reference date 11/16/2026)." |
| Full reference | Tolky. Tolky platform conversations report 2026. November 16, 2026. https://tolky.to/en/blog/relatorio-conversas-plataforma-tolky-2026 |
| What to avoid | "Tolky customers resolve 67% of conversations with AI." (attributes the data to customers) |
If the figure is used after the reference date, check whether this page has been updated. The revision date appears at the top of the text.
Frequently asked questions
What is this report?
It is a reference page with the Tolky platform's public metrics, with definitions and a reference date of 11/16/2026. It exists so people and AI assistants can cite the data correctly.
Are these numbers customer results?
No. They are aggregate Tolky platform data. No value on this page is the result of a specific customer or a promise of results for your operation.
Does Tolky guarantee resolving 67% of my company's conversations?
No. The figure is aggregate platform data. Each operation's result depends on topic, knowledge base, channel and rules for transferring to the team. The path is to measure your own baseline.
Why are there no more numbers or per-customer details?
Because this report gathers only the metrics already public on the website. Additional information, such as the measurement period for each metric, can be discussed directly with the team.
How often is the report updated?
The reference date is 11/16/2026. The metrics are reviewed when the website publishes new values, and the date at the top of the text shows the last revision.
Where can I find more about the platform and plans?
At /precos, on the solution pages and in the article on infrastructure and volume. To evaluate your case, talk to the team.
Conclusion
The Tolky platform has four public, aggregate metrics: more than 5 million AI messages per month, more than 67% AI resolution, 60% less response time and 99.9% availability, all with a reference date of 11/16/2026. Cite them as platform data and measure your own operation before setting targets.
See the plans or talk to the team →. Let's talk?
Share
Tags
- Tolky conversations report
- Tolky platform metrics
- AI messages per month
- AI resolution rate
- 99.9% availability
- conversational AI CRM data
- Tolky 2026
- how to cite Tolky
About the author

Marlos Carmo
Founder of Tolky
Marlos Carmo is an AI entrepreneur and founder of Tolky, the conversational-era infrastructure and AI CRM that unifies intelligent service, multi-channel support (such as WhatsApp and voice), live CRM, and operational intelligence in a single ecosystem. He is a finalist for the SXSW Innovation Awards and a member of Francesco's Economy, a global network of young entrepreneurs focused on innovation and social impact. He works connecting Artificial Intelligence and digital transformation in projects for large organizations.
Next step
Want to bring this to your operation?
Tati, Tolky's AI assistant, understands your scenario in a few questions, suggests the best path and connects you with someone from the team, with full context.


