The goal of Google Search has always been simple: to help you ask anything on your mind. Whether it is a quick fact to help with your daily hustle or a complex question about starting a new business, Nigerians rely on Search every single day.

Over the last year, Google has rapidly reimagined what Search can do with AI. The momentum has been incredible—just one year after its debut, AI Mode has surpassed one billion monthly users globally. As people have realised just how much more Search can do for them, they are searching more than ever before, reaching an all-time high in search queries last quarter. Today at Google I/O, Google shared the next step in its journey to bring together the best of a search engine with the best of AI.

To power this next chapter, Google is officially upgrading Search with Gemini 3.5 Flash as the new default model in AI Mode for everyone worldwide. Delivering sustained frontier performance for agents and coding, Gemini 3.5 Flash is the engine driving the new era of AI-powered Search. Because curiosity doesn’t always fit into standard keywords, this powerful AI model is transforming Search from a tool that simply finds information into an intelligent platform capable of reasoning, monitoring the web, and executing complex tasks on your behalf.

Here is a look at the four biggest AI-powered announcements coming to Google Search:

1. A Completely Reimagined Search Box

Google is introducing the biggest upgrade to its Search box in over 25 years. Now completely reimagined with AI, the new intelligent Search box dynamically expands to give you the space to describe exactly what you need. It goes beyond simple autocomplete by anticipating your intent and helping you phrase your questions. You are no longer limited to typing; you can now search using text, images, files, videos, or even Chrome tabs as inputs. Additionally, Google is making it easier to ask follow-up questions directly from an AI Overview, flowing naturally into a conversational back-and-forth where your context stays with you as you explore.

2. New Search Agents That Work in the Background

We are entering the era of Search agents, where you can create and manage multiple AI agents directly in Search. Google is launching “Information agents” that operate in the background 24/7. These agents intelligently scan the web—alongside fresh data on finance, shopping, and sports—to monitor for changes related to your specific questions. For example, if you are house hunting, your agent will continuously scan the market and notify you the moment a listing matches your exact criteria. Furthermore, Search is expanding its agentic booking capabilities; you can soon share specific criteria (like a late-night private karaoke room) and Search will pull the latest pricing and links to finish booking. For certain categories, Google can even call businesses on your behalf.

3. Custom Mini-Apps and Visuals Built Just for You

Search is no longer just returning links; it is now building the ideal response in the perfect format for your query entirely on the fly. By bringing the power of Google Antigravity and the agentic coding capabilities of Gemini 3.5 Flash into Search, users will get a custom “Generative UI.” This means Search can design custom layouts, interactive visuals, tables, graphs, or simulations in real-time. But it goes a step further: if you have an ongoing task, like establishing a new health routine, Search can actually code a custom fitness tracker or mini-app for you. These custom dashboards tap into real-time sources like live maps and weather, giving you a personalised tracker you can return to again and again.

4. Expanded Personal Intelligence Without a Subscription

For AI to be truly helpful, it shouldn’t just know the world’s information—it should understand your personal context, too. To achieve this, Google is expanding Personal Intelligence in AI Mode to more people in nearly 200 countries and territories across 98 languages. Crucially, this is being rolled out with no subscription required. Users can securely connect apps like Gmail, Google Photos, and soon Google Calendar directly to Search. Designed with transparency and choice at its heart, this allows you to safely ask Search to find information buried in your own personal files, always keeping you in complete control of your connected data.

By Modupe Gbadeyanka

The chief executive of Dimensions Data Limited, Mr Gbenga Olabiyi, has blamed road construction for 60 per cent of network outages caused by fibre cuts.

Speaking recently at the National Dig-Once Policy Forum, which marked the 8th Policy Implementation Assisted Forum (PIAFo), he drew attention to the gap between the infrastructure Nigeria has and what it can actually deliver if a coordinated framework is adopted.

“Nigeria currently has about 35,000 kilometres of fibre in the ground, yet only 16 per cent of Nigerians are connected to it. Broadband penetration stands at 45 per cent. Lagos alone has a penetration rate of over 70 per cent,” Mr Olabiyi said.

He emphasised that the failure to address the missing fibre link over the years has led to saturation of connectivity in urban centres, while the hinterlands are left either unconnected or poorly served.

At the same programme, convened by Mr Omobayo Azeez, stakeholders in the telecommunications sector called for the adoption of the dig-once policy to lower the costs of fibre deployment, reduce infrastructure damage, improve safety, and shorten rollout timelines.

Quoting the Nigerian Communications Commission (NCC), it was noted that of the 50,000 fibre cut incidents recorded in a year, about 30,000, which represents 60 per cent, occurred during road construction and rehabilitation.

Stakeholders thus called for a review of existing road construction and building codes to accommodate the installation of fibre conduits in the original design standard of the infrastructure planning.

“What Dig-Once offers is an opportunity to correct this,” the president of the Association of Telecommunication Companies of Nigeria, Mr Tony Emoekpere, stated.

He added that even operators frequently damage one another’s cables during repeated digging, thus increasing repair costs and service disruptions.

The Deputy Director of Strategic Business Initiatives at ipNX Nigeria Limited, Mr Segun Okuneye, said under the dig-once policy, road contractors should install ducts during construction.

He said the repeated excavation of the road leads to incessant destruction of existing infrastructure and triggers service blackouts with operators bearing additional costs of repair of replacing the fibre.

Also, the chairman of the Association of Licensed Telecom Operators of Nigeria (ALTON), Mr Gbenga Adebayo, said operators should focus not just on digging once but on eliminating unnecessary digging altogether by sharing existing infrastructure and jointly replacing legacy cables.

“Early fibres laid 15 to 20 years ago are now ageing, and the industry needs a plan to replace them without everyone digging the same routes again,” he said.

One of the most powerful ways to modernize a support team is by optimizing support operations with AI automation.

When implemented thoughtfully, automation doesn’t replace human agents—it elevates them.

Less duplicated work‚ resolving issues faster‚ and providing a consistent experience on different channels․

Since there are many content formats in the current content environment‚ the most effective formats for guidelines are those that are structured‚ practical‚ and focused on “what you can actually do,”‚ as opposed to abstract theory․

This article follows that same intent‚ and seeks to document the journey to better customer support automation‚ without mentioning brand names or links․

1. Automate first‑contact triage with smart workflows

One of the common entry points into support automation is to respond to end customers’ requests as soon as they are received‚ rather than making them wait for a human agent to pick up a chat channel or email, as is customary․

This is done through clever workflows that ask a few questions‚ qualify and classify the issue, and recommend a next best action․

This can mean transferring to a human agent‚ directing the user to a specific help article‚ or beginning a guided self-service flow right within the chat․

It reduces the friction to get started and shows customers they are being heard from the first message․

It reduces the burden on agents because they only see tickets that require human judgement․

In the best implementations‚ the bot feels like a helpful assistant and not a hindrance to the customer reaching a resolution․

2. Route tickets with intelligence, not just speed

Responsiveness is still important‚ but smart routing is what takes automation from simple ticketing to responsive‚ scalable support․

Instead of shuffling tickets to the agent available‚ the system can route based on the subject‚ difficulty‚ language‚ or even expected resolution․

This way, billing problems are routed to billing experts‚ complaints about product setup are routed to technical experts‚ and routine status inquiries are routed to agents who can handle volume․

For example‚ clever routing could send high-touch or high-stakes tickets to a more senior agent or tickets that ask the same question repeatedly to agents specialized in a specific workflow․

It’s this kind of intelligence that allows teams to be faster and happier when they are thinking of routing-based automation beyond simple round robin distribution․

3. Turn FAQs into self‑service journeys

Another area the current content focuses on is changing the format of FAQs into more interactive self-service experiences that guide customers through flows‚ checklists, or conversational search‚ as opposed to serving them a long list of links‚ making it easier to discover a solution․

This reduces the need to create tickets in the first instance and reduces the support workload by focusing on more high-value ‚ complex interactions․

An organized help center can examine common patterns in failed searches and proactively suggest the most appropriate articles or troubleshooting steps․

It can also serve as the backbone for mini-chatbots that can guide the user through setup‚ configuration‚ or troubleshooting paths without opening a ticket․

When teams invest in AI automation to better support their operation‚ a self-service capability is often one of the first investments made․

4. Automate routine follow‑ups and escalations

In modern ticketing systems‚ the entire life cycle of the ticket from its creation can be automated․

Instead of relying on agents to remember to do a status update‚ a satisfaction survey‚ or an escalation‚ rules can be set up to automate these processes․

For example‚ if a ticket has been placed in the “pending customer reply” state for a specified period of time‚ a notification to the customer can be sent out to remind them‚ or if a complaint has not been resolved within any specified period‚ the ticket can be escalated to a manager․

The result is processes that are always followed‚ never missed SLAs‚ less manual work‚ and agents are freed up from the tedious tracking of time and sending reminders to focus on resolving problems․

A mix of automation and human intervention is often the optimal solution for a better experience for customers and agents alike․

5. Use AI to draft and summarize responses

AI-assisted writing has become the norm to scale support teams‚ with the tool helping staff draft an initial response‚ summarize long email threads‚ and suggest templated replies which agents personalize․

It is especially useful in high-volume or multi-language support environments‚ where replies to common questions must be timely and consistent․

This type of automation doesn’t replace agents‚ but acts as a force multiplier for them‚ ensuring that baseline questions are answered correctly and on-brand‚ while still enabling subtlety and empathy in less obvious situations․

Some teams use AI to translate or simplify support communications for different audiences‚ enabling them to support global customers without needing to hire additional staff․

6. Automate onboarding and welcome communications

Automation can also play an important role in onboarding‚ by providing an automated welcome sequence for new customers to help them get set up‚ implement best practices‚ and learn about key features and resources․

These sequences can incorporate email‚ in-app messages‚ and chat prompts to create a cross-channel experience․

To the support staff who deal with these customers‚ this reduces the number of “I don’t know where to start” help desk questions that pile up in the first few days after signing up․

Perhaps more considerably‚ walking users through the most important workflows has contributed to increased activation and retention rates․

One of the most visible ways to use AI automation in support is by transitioning from firefighting to empowering customers and agents with self-service and insights․

7. Trigger proactive support with behavior signals

An even more advanced form of automation involves proactively reaching out to customers before they reach out to you by identifying usage trends or risk signals based on the way they are using the product․

For example‚ when a user repeats the same action‚ fails to complete a key workflow‚ or is beginning to disengage‚ a system could send a personalized message or offer assistance before the customer churns․

These models may be based on behavioral analytics and artificial intelligence models‚ which have tracked tens of thousands of data points‚ events‚ and user behaviors to identify signals that can be used in a support flow to prevent and surface issues before they arise to improve customer satisfaction․

As well‚ proactive messaging must be finely tuned so as not to be perceived as spam‚ and teams iterate based on feedback and response rates․

8. Automate feedback collection and analysis

Many teams capture this feedback automatically as part of their improvement processes‚ for example‚ automatically sending out a customer satisfaction survey once a ticket is closed or analyzing customer messages to understand the sentiment․

This can also support testing‚ benchmarking‚ tracking performance versus targets‚ identifying trends and patterns to tackle, and prioritizing product or process changes․

For support leaders‚ this automation means raw interaction data is transformed into structured insights․

Instead of manually reviewing tickets‚ they view dashboards containing information about common problems‚ emerging topics‚ agent performance‚ etc․

Another effective way to improve efficiency in support is through AI automation․

Every interaction can be a learning and improvement opportunity․

9. Integrate omnichannel experiences

Omnichannel integration is a common thread in customer support automation workflows․

Customers do not care what channel they are in․

Customers expect the context to move with them as they continue the conversation via chat‚ email‚ phone‚ social media, or in an in-app message․

Automation across channels offers the advantage that each interaction builds on prior interactions‚ instead of beginning with a blank slate․

For example‚ if a customer starts a chat conversation and later sends an email‚ we want to show the chat conversation in the history view for the email conversation‚ and vice versa‚ so that the agent doesn’t have to ask the customer for context each time․

This is a feature that differentiates fragmented support experiences from single-threaded experiences‚ and is a common area of focus for teams modernizing their support workflows․

10. Build a feedback‑driven automation roadmap

The best customer support automation is not a single project․

Top teams start by identifying manual activities that take the most time or happen most often‚ and then determine which of those can be fully or partially automated․

They roll out gradual changes and analyze their effects in order to improve them based on real-world data․

This roadmap often includes:

• Pinpointing the top 20% of support scenarios that consume the most time.
• Designing workflows that combine bots, knowledge bases, and human agents.
• Continuously monitoring metrics like resolution time, satisfaction scores, and agent workload.

By combining this with a full focus on AI automating support operations‚ your support function can be scaled better․

There are options emerging like Ferndesk, which do seem to align with most of these points․

But the fundamental principle remains for any support teams the same: to automate support to be faster‚ smarter, and more human․