Skip to content

FAQ / Answers

Frequently asked questions about working with NEX4R.

Straight answers to the questions buyers ask about AI automation, AI agents, custom software and how NEX4R works. If your question is not here, email info@nex4r.com.

FAQ

General questions

What is NEX4R?

NEX4R is a technology brand of JDE Business Solutions (Thailand) Co., Ltd. that designs and builds AI automation, AI agents, custom AI applications, websites and custom software, and provides AI training and consulting. It serves businesses in Thailand, the Philippines, Malaysia, Singapore and the UAE, and clients worldwide online.

What is the difference between AI automation and an AI agent?

AI automation runs a defined process, using rules for predictable steps and AI for reading documents or messages. An AI agent is given a goal and decides the steps itself, using tools in your systems. Automation suits stable processes; agents suit multi-step tasks that vary from case to case.

Do we need to replace our current software?

Usually not. Most of our work connects the CRM, accounting, e-commerce, messaging and office tools you already use through their APIs. If a tool cannot be connected, we explain the options before building anything.

Can you build systems that work in Thai and other local languages?

Yes. We build websites, assistants and document processing for English together with Thai, Filipino, Malay, Chinese and Arabic, and we test AI features on content in each language rather than assuming they work.

Which AI models do you use?

We choose models per task, weighing quality, speed, cost, language support and where data is processed. We can work with models from providers such as OpenAI, Anthropic and Google, and we design systems so the model can be changed later.

Is our data used to train AI models?

We use model providers' business or API terms, under which customer data is generally not used for training, and we confirm the terms of the specific provider during design. Where data must not leave your environment, we tell you which options fit.

How long does a typical project take?

It depends on scope. Because we deliver in phases, a first working version is often possible in weeks, not months, with later phases adding more. The proposal sets out each phase and what it delivers.

How do you keep AI systems accurate and safe?

We ground AI answers in your own content, test on your real examples before launch, route uncertain cases to a person, require approval for important actions and log every run. Accuracy is measured and reviewed, not promised.

Who owns the software and data?

Your data is yours. Ownership of code and accounts is set out in the project agreement; our usual approach is that the client owns the software built for them and the cloud accounts it runs in.

Do you provide support after launch?

Yes, if you want it. We can monitor, maintain and extend systems under an agreed arrangement, or hand over documentation and train your team to run them.

Can we work with NEX4R remotely?

Yes. Discovery calls, design reviews, testing and training all work online, and we schedule across time zones from UTC+4 in the UAE to UTC+8 in the Philippines, Malaysia and Singapore. In-person meetings can be discussed where our teams are.

How do we start?

Send a short description of what you want to automate or build through the contact form or to info@nex4r.com. We reply by email, arrange a discovery call and then send a written proposal.

Questions by topic

Services

Questions about AI Automation8 questions
What is AI automation in simple terms?

AI automation is software that carries out repeated business tasks for you, using fixed rules for predictable steps and an AI model for steps that need reading or judging text. A person still owns the process and handles exceptions. It is different from a chatbot: the goal is finished work, not a conversation.

Which business processes are best to automate first?

Start with work that is frequent, rule-based and annoying: approvals, data copied between systems, reminders, document intake and recurring reports. These have clear triggers and clear outcomes, so it is easy to check the automation is right. Processes that change every week or depend on one person's judgement are better left for later.

How long does it take to automate a workflow?

It depends on scope and on how many systems are involved. A single workflow connecting two or three tools is often working in weeks, not months; a multi-department programme is delivered in stages. We give you a written plan with phases after the discovery call.

Do we have to replace our existing software?

No. AI automation connects the tools you already use through their APIs, such as your CRM, accounting software, shop and messaging channels. If a tool has no API, we tell you the options before building anything.

What happens when the automation gets something wrong?

Every workflow has defined exception paths: if a rule does not match or the AI model is not confident, the item goes to a named person instead of being guessed. Each run is logged, failures raise an alert, and there is always a manual override.

Is AI automation PDPA compliant in Thailand?

Automation can be built to support compliance with Thailand's Personal Data Protection Act B.E. 2562 (2019), but compliance depends on how your business collects and uses the data, not on the software alone. We design for data minimisation, access control and logging, and we tell you where personal data flows so your data protection lead can review it.

What do we need to provide to get started?

A person who knows the process, access to the systems involved (test accounts first, where possible), examples of real inputs such as invoices or emails, and the rules your team follows today. We document everything else during discovery.

Who maintains the automation after launch?

You choose. We can hand over documentation and train your team to maintain it, or support and change it for you under an agreed arrangement. Either way, you own the workflows and the accounts they run on.

Explore AI Automation
Questions about Agentic AI7 questions
What is the difference between an AI agent and a chatbot?

A chatbot replies to messages. An AI agent works towards a goal: it can use tools, take several steps, check results and act in other systems. That makes agents useful for multi-step work, and it is also why they need clear permissions and approval gates.

How long does it take to build an AI agent?

It depends on the task and on how many systems the agent must use. A focused agent with a few tools can often reach a working pilot in weeks, not months; broader agents are built in stages. We give you a phased plan after discovery.

Are AI agents fully autonomous?

Only as autonomous as you decide. We set autonomy per action: reading and drafting can run freely, while sending, paying, deleting or changing customer data can require approval. Limits are relaxed only when the logs show the agent is reliable.

How do you stop an AI agent from making mistakes?

No AI system is error-free, so we design for safe failure. Agents get narrow tools rather than broad access, outputs are validated, uncertain cases go to a person, and a test set is run before every release. Every action is logged so mistakes are visible and traceable.

Which AI models do you use for agents?

We choose the model per task, weighing reasoning quality, speed, cost, language support (including Thai) and where data is processed. We can work with models from providers such as OpenAI, Anthropic and Google, and can design so the model can be changed later.

Can we see why an agent made a decision?

Yes. Each run is recorded with its inputs, the tools it called, what those tools returned and the final output. Your team can review runs in a log or dashboard, which is also how quality is checked over time.

Is our data used to train the AI model?

We use model providers' business or API terms, under which customer data is generally not used for training, and we confirm the terms for the specific provider during design. Where data must stay in a particular place or not leave your systems, we tell you which options fit.

Explore Agentic AI
Questions about Custom AI Development7 questions
What is RAG in AI?

RAG, or retrieval-augmented generation, is a method where the system first searches your own documents for relevant passages and then asks the language model to answer using only those passages. It keeps answers tied to your content and lets the system show its sources. It does not remove the need for testing, because retrieval can still miss or pick the wrong passage.

How much data do we need for a custom AI application?

Usually less than people expect. Most business applications use an existing model with your documents and examples, not a model trained from scratch. What matters is that the content is current, reasonably organised and that you can give us real questions to test against.

Do you train our own AI model?

Rarely, and only when it is clearly better. Most needs are met by combining an existing model with retrieval, good instructions and tools. Fine-tuning can help for narrow, repetitive tasks; we will tell you if your case is one of them and what it would involve.

Can a custom AI system work in Thai?

Yes. Current large language models handle Thai, and we test retrieval and answers on Thai content specifically, because word segmentation and mixed Thai-English text need care. The same applies to Filipino, Malay, Chinese and Arabic content.

How do you make AI answers accurate?

We ground answers in your documents, ask the model to say when it does not know, test on a set of your real questions before launch and after every change, and route uncertain cases to a person. Accuracy is measured and reported, not promised.

Where is our data stored and processed?

That is a design decision we make with you. Content can be stored in your own cloud account, and model providers and hosting regions are chosen with your data protection obligations in mind, such as Thailand's PDPA or Singapore's PDPA. We document every place data goes.

Who owns the software and the data?

Your data stays yours. Ownership of the code and the accounts it runs on is set out in the project agreement; our usual approach is that the client owns the application built for them and the cloud accounts it runs in.

Explore Custom AI Development
Questions about Website Development7 questions
How long does it take to build a business website?

It depends on the number of pages, languages and features, and on how quickly content is ready. A focused company site can often launch in weeks, not months; larger sites and web applications are delivered in phases. We give you a written plan after discovery.

What technology do you build websites with?

We usually build with modern frameworks such as Next.js and React, a relational database such as PostgreSQL and a CMS suited to your team. We choose the stack for speed, security and how easy it is for you to maintain, and we document it.

Will my website be optimised for SEO?

Yes. Technical SEO is part of the build: page structure, metadata, structured data, sitemaps, internal linking, page speed and redirects from old URLs. Rankings also depend on content and competition, so we do not promise positions.

Can you build a website in Thai and English?

Yes. We build multilingual sites with a separate URL for each language, hreflang tags and a CMS where each language is edited side by side. We also support Filipino, Malay, Chinese and Arabic, including right-to-left layout for Arabic.

Can our team update the website ourselves?

Yes. Pages, articles, FAQs and products are edited through an admin with user roles, so your team can publish without a developer. We train the people who will use it.

Do you redesign existing websites without losing rankings?

We plan a rebuild to protect search visibility: we map every old URL to its new one, keep valuable content, set up permanent redirects and monitor Search Console after launch. Some movement in rankings after a relaunch is normal; the redirect plan keeps it small.

Do you provide hosting and maintenance?

We can deploy to hosting you own or recommend options, and we can support and update the site after launch under an agreed arrangement. You keep ownership of the domain and hosting accounts.

Explore Website Development
Questions about SEO Automation7 questions
What is SEO automation?

SEO automation is using software to perform repeatable search-optimisation tasks such as technical checks, structured data, metadata suggestions and reporting. It saves time and catches problems earlier. Strategy, writing and final approval remain human work.

Does SEO automation replace an SEO specialist?

No. It removes the repetitive checking and reporting so a specialist or marketer can spend time on strategy, content and links. Someone still needs to decide priorities and review what the automation suggests.

Will AI-generated content hurt my Google rankings?

Google's guidance focuses on whether content is helpful and accurate, not on how it was produced. Mass-produced, unreviewed AI text is risky because it tends to be generic and sometimes wrong. We use AI for research, briefs and drafts, with a person editing and approving everything that publishes.

What is GEO and how is it different from SEO?

GEO, or generative engine optimisation, is making content easy for AI answer engines to understand and quote. It builds on SEO: clear answers near the top of a page, well-structured headings, FAQs, structured data and named entities. Good SEO foundations remain the starting point.

How often is technical SEO checked?

As often as your site changes. Most sites benefit from weekly crawls, with daily checks on key pages and alerts for sudden changes such as pages dropping out of the index. We agree the schedule with you.

Can you automate SEO for a multilingual site?

Yes. Automation is especially useful there: it can check that each language version exists, that hreflang links are correct and that metadata is translated rather than copied. We support English with Thai, Filipino, Malay, Chinese and Arabic.

How long before SEO automation shows results?

Technical monitoring and reporting are useful from the first run. Ranking changes depend on what is fixed, your content and your competitors, and usually take longer. We do not promise rankings or traffic numbers.

Explore SEO Automation
Questions about AI Training & Consulting7 questions
Who is AI training for?

It is for any team that already uses, or plans to use, AI tools at work, from front-line staff to senior managers. We tailor content to each audience: hands-on practice for staff, decision-making and risk for leaders.

Do you run AI training in Thai?

Sessions are designed in English, and Thai-language delivery can be discussed for teams in Thailand. Examples and exercises can use your own documents, including Thai-language material. Tell us your team's languages when you get in touch.

Is the training online or in person?

Both are possible. Teams near our people in Thailand, the Philippines, Malaysia and Singapore can have in-person sessions, and every programme can run online for teams anywhere, including the UAE.

Do we need to be a NEX4R client to book training?

No. Training and consulting can be booked on their own. Some companies use them first to decide whether to build anything; others never do, and that is fine.

Will you help us write an AI usage policy?

Yes. We draft a short, practical policy covering approved tools, what data may and may not be shared, human review and accountability. It should be reviewed by your own legal or data protection lead before adoption, particularly against laws such as Thailand's PDPA.

How long is an AI workshop?

Formats range from a short briefing to a full-day workshop or a series of sessions. We recommend a format once we know your audience and goals.

Which AI tools do you teach?

We teach the tools your company already has or plans to approve, such as ChatGPT, Claude, Gemini or Microsoft Copilot. The habits we teach, like giving context, checking sources and protecting data, carry over between tools.

Explore AI Training & Consulting
Questions by topic

Products

Questions about NEX4R AI Assistant7 questions
What is NEX4R AI Assistant?

NEX4R AI Assistant is a business AI assistant configured for each company. It answers from that company's own documents and data, can perform permitted actions in connected systems and hands conversations to staff when needed.

Can the AI assistant work on LINE and WhatsApp?

Yes. It can be connected to a LINE Official Account, the WhatsApp Business Platform, Facebook Messenger and Viber, as well as your website. Each channel has its own setup and account requirements, which we handle with you.

What data can the assistant access?

Only the sources you connect and the actions you allow. Access is set per source and per action, and sensitive systems can be left out entirely. Conversation logs are kept so you can see what it was asked and how it answered.

What happens when the assistant does not know the answer?

It is instructed to say it does not know rather than guess, and to offer a person. Hand-over rules can also trigger on topics such as complaints, refunds or legal questions.

Does it answer in Thai?

Yes. The assistant replies in the customer's language, including Thai, and we test it on your Thai content and real Thai questions before launch. The same applies to Filipino, Malay, Chinese and Arabic.

How long does it take to set up?

It depends on the amount of content, the channels and any system actions. A website or single-channel assistant answering from existing documents is often live in weeks, not months; actions and multiple channels add time.

Is a customer-facing AI assistant compliant with data protection law?

It can be set up to support compliance: a clear notice to users, collecting only what is needed, access controls and retention rules. Your obligations depend on your business and market, for example under Thailand's PDPA or the Philippines Data Privacy Act, and should be reviewed by your data protection lead.

Explore NEX4R AI Assistant
Questions about NEX4R Lead Engine6 questions
What is NEX4R Lead Engine?

NEX4R Lead Engine is a configured system that captures leads from all your channels, scores them against your criteria and routes them into your CRM and to the right salesperson. It is set up for each business rather than sold as a fixed template.

How does lead scoring work?

Scoring combines rules your sales team agrees, such as country, company type, budget or timeline, with AI reading the free-text message to judge intent and fit. Scores are explained in plain words so sales can see why a lead ranked where it did.

Does it replace our CRM?

No. It works with the CRM you already use, such as HubSpot, Salesforce, Zoho CRM or Pipedrive, creating and updating records there. If you do not have a CRM yet, we can help you choose one.

Does it send emails to leads automatically?

Only if you want it to and only with templates your team has approved. Many companies start with drafts that a salesperson reviews and sends. Outreach follows the consent and opt-out rules of the markets you sell in.

Is automated lead handling allowed under data protection laws?

It can be, when personal data is collected with a clear purpose, kept to what is needed and handled with the rights of the person in mind. Laws differ by market, for example Thailand's PDPA, Singapore's PDPA and Malaysia's PDPA 2010; we design the flows so your team can meet them and recommend a review by your data protection lead.

How long does it take to set up?

It depends on the number of sources and the CRM setup. Connecting a website form, one or two ad sources and a CRM with basic scoring is often working in weeks, not months; more channels and markets are added in stages.

Explore NEX4R Lead Engine
Questions about NEX4R CMS6 questions
What is NEX4R CMS?

NEX4R CMS is a content management system that NEX4R sets up with the websites it builds. Editors manage pages, articles, products and FAQs in several languages, with SEO fields, roles, version history and an API.

Can non-technical staff use it?

Yes. Editing happens in plain forms with labels and help text, not in code. We train your editors and give them a short guide.

Does the CMS support Thai and other languages?

Yes. Each piece of content has a version per language with its own URL and SEO fields, and the site outputs hreflang so search engines show the right version. Thai, Filipino, Malay, Chinese and Arabic content can be managed alongside English.

Is NEX4R CMS headless?

It works both ways: it powers the website directly, and content is also available through an API for apps, other sites or an AI assistant.

Can we move our existing website content into it?

Yes. Content migration is part of setup. We map old pages to new content types, move the text and images, and set permanent redirects from old URLs to protect search visibility.

How is the CMS kept secure?

Sign-in supports two-factor authentication, permissions are checked on the server for every action, changes are recorded in an audit log and the public site does not expose the admin. We keep dependencies updated as part of support.

Explore NEX4R CMS
Questions about NEX4R AI Trader6 questions
What is NEX4R AI Trader?

NEX4R AI Trader is software infrastructure for running, monitoring and reviewing trading strategies that the client defines. It handles execution, risk limits, data, logs and alerts. It is not investment advice, a signal service or a managed account.

Does NEX4R AI Trader guarantee profits?

No. No software, AI model or backtest can guarantee trading results, and you can lose money, including more than you expect in fast markets. The product helps run a strategy consistently and within limits; whether the strategy makes or loses money is outside our control.

Does NEX4R give trading or investment advice?

No. NEX4R provides technology. Strategy choices, instruments and position sizes are the client's decisions, and clients should take independent professional advice where appropriate.

How is risk managed?

Limits on position size, exposure and daily loss are enforced by the system, with an automatic stop when they are reached and a manual kill switch. Monitoring and alerts tell named people when something fails. These controls reduce some operational risks; they do not remove market risk.

Can we test a strategy before trading live?

Yes. Strategy logic can be tested against historical data and on a demo or paper account first. Backtests and demo results have limits, such as slippage and changing market conditions, and do not predict live performance.

Who is responsible for regulatory compliance?

The client is responsible for the trading activity and for any licences or rules that apply in their jurisdiction. If you operate a regulated business or trade for others, check your obligations with your regulator or legal adviser before using any automated trading system.

Explore NEX4R AI Trader
Questions about Custom SaaS7 questions
What is custom SaaS development?

Custom SaaS development is building web-based software designed for one company's workflows or for a product that company will sell, instead of adapting a packaged tool. It is delivered through the browser, runs in the cloud and is improved continuously.

How long does it take to build a custom SaaS platform?

It depends on scope. We plan in phases so a first working version that solves the main problem is often possible in weeks, not months, with later phases adding features. You get a written phase plan before work starts.

Who owns the software you build?

Ownership is set out in the project agreement. Our usual approach is that the client owns the application built for them, its data and the cloud accounts it runs in.

Should we build custom software or buy an existing product?

Buy when an existing product fits most of your process at an acceptable cost. Build when the process is central to how you compete, when several tools are being stitched together or when you plan to sell the software. We will tell you honestly if a packaged product would serve you better.

Can custom SaaS integrate with our accounting and CRM systems?

Yes, where those systems offer an API, which most modern ones do, such as Xero, HubSpot and Google Workspace. We confirm each integration during discovery.

How do you handle security and data protection?

We build with secure sign-in, server-side permission checks, encryption in transit, backups and audit logs, and we design data handling with laws such as Thailand's PDPA, Singapore's PDPA and the UAE PDPL in mind. Security is reviewed again before each release.

What happens after launch?

We can support, monitor and extend the platform under an agreed arrangement, or hand over code and documentation to your own team.

Explore Custom SaaS
Questions by topic

Solutions

Questions about Business Automation6 questions
What is business process automation?

Business process automation is using software to carry out the routine steps of a business process, such as moving data, routing approvals and sending notifications, so people only handle decisions and exceptions. With AI added, it can also read documents and messages that used to need a person.

Which departments should we automate first?

Usually finance and administration, because their processes are frequent, rule-based and easy to check. The best first process is one that takes real time every week and has clear rules. We help you rank the candidates.

How disruptive is it to automate a business process?

We keep disruption low by automating one process at a time, testing on real examples and running the automation alongside the manual process before switching. Staff are trained before go-live and keep a manual override.

Can automation keep an audit trail for compliance?

Yes. Every automated step and every approval can be logged with time and person, which often gives a clearer record than email-based processes. Retention of those records is set to match your policies.

Do we need new software to automate our processes?

Usually not. Most automation connects the accounting, HR, CRM and office tools you already use. If a tool cannot be connected, we explain the options and their trade-offs.

How long does business automation take?

The first process is often working in weeks, not months, depending on the systems involved. A wider programme is delivered process by process, so value starts before the whole programme is finished.

Explore Business Automation
Questions about Sales & Marketing6 questions
How can AI help a sales team?

AI can read and qualify incoming enquiries, draft first replies and follow-ups, summarise accounts before a call and keep CRM records tidy. The salesperson still owns the relationship and approves what is sent. The main gains are speed and consistency.

Do we need to change our CRM?

No. We work with the CRM you use, such as HubSpot, Salesforce or Zoho CRM. If you do not have one, we can help you choose and set it up.

Can you connect LINE and WhatsApp leads to our CRM?

Yes. Conversations from a LINE Official Account, the WhatsApp Business Platform and Facebook Messenger can create or update contacts and deals in your CRM, so chat leads are tracked like any other.

Is automated marketing follow-up legal?

It can be, when people have agreed to hear from you or there is another lawful basis, messages identify you and opting out is easy. Rules differ by country, for example under Thailand's PDPA or Singapore's PDPA and its Do Not Call provisions, so we design consent and opt-out handling into every sequence.

Will AI write our marketing content?

AI helps with research, briefs, outlines, first drafts and translations. A person edits, checks facts and approves everything before it is published, because generic or inaccurate content harms both trust and search visibility.

How do we know which campaigns work?

By recording source and campaign on every lead and carrying it through to deals in the CRM. Reports then show leads, qualified leads and closed deals per campaign, not just clicks.

Explore Sales & Marketing
Questions about Customer Support6 questions
Will AI replace our customer support team?

No. AI takes the repeated questions and helps agents with drafts and context; people handle complaints, judgement calls and anything sensitive. Most teams use the time saved to answer complex cases faster and better.

How accurate are AI support answers?

Accuracy depends on the quality of the knowledge base and on testing. We ground answers in your approved content, test on real past questions before launch, tell the assistant to admit when it does not know and review logs regularly. No AI system is perfect, which is why escalation is built in.

Can AI support work on LINE and WhatsApp?

Yes. The assistant can work on a LINE Official Account, the WhatsApp Business Platform and Facebook Messenger, as well as web chat and email, with hand-over to staff on each channel.

Does it integrate with our helpdesk?

Yes, where the helpdesk has an API, which tools such as Zendesk, Freshdesk and HubSpot Service Hub do. The assistant works inside your existing setup rather than replacing it.

How are complaints and sensitive cases handled?

They go to a person. We agree the topics and signals, such as complaints, refunds, legal threats or distress, that trigger immediate hand-over, and the agent receives the full conversation.

How is customer data protected?

Customers are told they are talking to an AI assistant, only needed data is used, access is controlled and logs are kept for a set period. We design with your obligations in mind, for example under Thailand's PDPA, the Philippines Data Privacy Act or the UAE PDPL.

Explore Customer Support
Questions about Operations6 questions
What kind of internal tools do you build?

Web-based tools for running daily operations: job and order tracking, scheduling, inventory and asset records, checklists, approvals and dashboards. Each is designed around a company's own workflow and works on desktop and phone.

Can you move our spreadsheets into a proper system?

Yes. Moving data from spreadsheets is a normal part of the work. We clean and map the data, import it and keep a read-only copy of the original for reference.

Should we build an internal tool or buy software?

Buy when a packaged product fits your process well. Build when your workflow is specific, when you are stitching several tools together or when the process is part of what makes you different. We will say so if an existing product is the better choice.

Can staff use the tools on their phones?

Yes. We design internal tools to work on phones as well as desktops, which matters for field, warehouse and site teams.

Can we get alerts on LINE or WhatsApp?

Yes. Alerts can go to LINE, WhatsApp, Microsoft Teams, Slack or email, depending on what your team uses and what each platform's business terms allow.

How long does it take to build an internal tool?

It depends on the workflow and integrations. A focused tool replacing one spreadsheet process is often in use in weeks, not months; broader platforms are built in phases.

Explore Operations

Your next advantage starts here

Let's build the system behind your next big move.

Tell us what slows your business down. We'll help you find the smarter way forward.

Start a conversation