
What happened
PKSHA Technology provided the conversational AI agent feature of its AI SaaS "PKSHA ChatAgent" to NTT Docomo's "Online Procedure Support" desk for contract and billing procedures.
Why it matters
The generative AI confirms a customer's intent by interacting with them, then narrows down the needed information while referring to procedure data and FAQs, so customers are not required to phrase their questions correctly.
What to watch
The design targets short inputs like "rate change" or "plan change," which are hard to pin down by intent, so whether the AI narrows intent accurately is the test.
WHO IT HITSThis lands on customer support and contact-center teams at telecom carriers, who handle complex contracts, rate plans and eligibility conditions. It may also affect teams building self-service online procedures, since the AI is meant to guide customers from vague input to the right step.
Summaries like this, every morning.
NTT Docomo customers do not always send fully formed questions. Short phrases like "rate change" or "plan change" make it hard to tell what the customer actually wants, and trying to be accurate often produces long explanations instead. In telecom services in particular, contract details, rate plans and eligibility conditions have grown complex, and online procedure options have multiplied, so a system that can guide customers to the right information even when they leave out the premises has become necessary.
Against that backdrop, PKSHA Technology provided the conversational AI agent feature of its AI SaaS "PKSHA ChatAgent" to NTT Docomo's "Online Procedure Support." The approach deliberately abandons the premise of making customers ask correctly. The generative AI talks with the customer, confirms the intent of the question, and narrows down the information while referring to procedure information and FAQs.
The test is likely to be whether the follow-up questions actually converge on the right intent when the input is only a few words. For support teams at carriers, the payoff hinges on whether this reduces the effort customers spend hunting for the right FAQ or procedure, rather than simply adding another layer of conversation.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
OpenAI published copy-ready prompts for GPT-6.2 Astra and Sol covering writing style, initiative, and coding…

Seramount's survey of more than 1,800 U.S

Cloudflare CEO Matthew Prince predicted the AI-driven security crunch will feel like the sky is falling for tw…

At the 73rd Listed Company Club on September 18, Bengoshi Dotcom CEO Taichiro Motoe traced an eight-year push…

AMD bought $2.5 billion of Anthropic stock at the $83 IPO price as part of a $5 billion commitment, and Anthro…

Levent Alpöge said he and Claude proved the Riemann Hypothesis and the Generalized Riemann Hypothesis; the Lea…
