It answers from your documents
Retrieval first, generation second. The model summarises what was found rather than recalling what it was trained on.
Freelance AI Chatbot Developer in Bangalore
A chatbot that answers from your content, not from its imagination.
Most business chatbots fail for one of two reasons: they answer from a decision tree nobody maintained, or they answer from a general model that cheerfully invents your refund policy. I build the third kind – one that reads your own documents and says so when it does not know.
- Answers grounded in your content
- Handover to a human built in
- Runs on your site or WhatsApp
- Replies within one working day
Overview
What a grounded chatbot actually is.
A language model on its own knows nothing about your business. Ask it about your delivery timelines and it will produce something plausible and wrong, because producing plausible text is the whole of what it does. That failure mode is not a bug to be prompted away; it is the technology working as designed.
The fix is retrieval. Your documents – product pages, policies, price lists, past support replies – are split, indexed and searched at the moment of each question. The model is handed the relevant passages and asked to answer from those alone. It is the difference between asking someone to recall your policy and handing them the policy to read aloud.
That single change moves the project from a novelty to something you can put in front of customers. It also makes the failure mode safe: when nothing relevant is found, the honest answer is “I do not have that” and a route to a person, rather than a confident fabrication that you find out about from an angry email.
What it does not do is replace your support team. It removes the thirty repeated questions a day that never needed a human, and it hands over everything else with the conversation attached. A chatbot sold as a replacement for staff is being oversold.
What I offer
Chatbotwork
01
Website chatbots
A widget on your site that answers from your own pages, documents and policies, in your tone rather than a vendor’s.
The questions that arrive at 11pm get answered at 11pm.
02
Document-grounded answers
Your PDFs, policy pages and product data indexed so answers are retrieved from them and cite where they came from.
You can check every answer against the source it used.
03
Lead capture and qualification
Conversations that collect what you actually need – requirement, budget range, timeline – before a human is involved.
Your inbox fills with qualified enquiries rather than “hi”.
04
WhatsApp and web, one brain
The same indexed content serving a web widget and a WhatsApp number, so both say the same thing.
One place to update when a policy changes.
05
Human handover
Clear escalation rules, with the whole conversation passed to whoever picks it up.
Nobody has to ask the customer to start again.
06
Fixing a chatbot that invents things
Auditing an existing bot, finding where it is answering ungrounded, and putting retrieval under it.
A bot you can stop apologising for.
Key benefits
What a properly built chatbot changes.
Six things that follow from grounding answers in your own content.
It admits what it does not know
No relevant passage means no answer and a route to a person – the single most important behaviour to get right.
It works outside office hours
The repeated questions get answered at night and at weekends, when they would otherwise wait.
It hands over cleanly
Escalation rules you set, with the full conversation attached so nothing is repeated.
It shows you what people ask
Logged questions are a content brief: the gaps in your site, written by your customers.
It stays correctable
Wrong answer? Fix the source document and reindex. You are editing content, not prompts.
Who this is for
Who a grounded chatbot suits.
And who is better served by something simpler.
Businesses answering the same thirty questions
Where support time goes on questions your own site already answers badly.
Stores with a real catalogue
Sizing, compatibility, stock and delivery questions that arrive before every order.
Service firms with documents
Policies, scope documents and price lists that customers will not read but will ask about.
Clinics and practices
Timings, preparation instructions and what to bring – asked constantly, rarely urgent.
Teams buried in WhatsApp
Where enquiries arrive on a personal number and nothing is logged or qualified.
Not for: five pages and no documents
With little content to retrieve from, a chatbot has nothing to ground on. A clear FAQ page is the honest answer.
Where I work
Why this is worth doing from Bangalore, for Indian businesses.
Indian customers ask before they buy, and they ask in the channel they already have open. For most businesses here that is WhatsApp rather than a web form, and the enquiry arrives as a voice note or a one-line message at an hour nobody is working. The question is rarely difficult – it is the same question as yesterday – but it still costs a person twenty minutes to find and answer.
Language is the part people underestimate. Customers write in English, in Hindi, in Kannada, and frequently in two at once, transliterated. A bot trained on a rigid script fails on the first message. A retrieval-based one copes far better, because the matching happens on meaning rather than on exact phrasing – though it is worth testing with your own customers’ real messages before anyone promises anything.
On cost: these systems are priced per conversation and per token, and those numbers are small until they are not. I model the running cost against your actual message volume before you commit, so the bill in month three is one you already saw in month zero.
Where customers already are
A WhatsApp number and a web widget answering from one source.
Mixed-language messages
Tested against how your customers actually write, not an idealised script.
Running cost modelled first
Per-conversation economics shown before you commit, not discovered later.
Indian working hours, or none
Handover routed to people who are awake, by rule rather than by hope.
Serving Chennai, Bangalore, India, Canada, UK, USA, Dubai, Singapore.
Industries
Where I have seen this earn its keep.
The pattern is the same: high question volume, low question variety.
E-commerce
Typically needs: Pre-purchase questions on sizing, compatibility, delivery and returns.
How I help: Catalogue and policy pages indexed, with order-status handover to a person.
Clinics & healthcare
Typically needs: Timings, preparation, documents to bring, what is covered.
How I help: Practice documents indexed, with anything clinical escalated immediately and by rule.
Education & training
Typically needs: Course content, fees, schedules, eligibility, placement questions.
How I help: Prospectus and course pages indexed, enquiries qualified before a counsellor calls.
Real estate
Typically needs: Availability, configuration, location, documentation.
How I help: Listing data and FAQs indexed, with site-visit requests captured properly.
Professional services
Typically needs: Scope, process, what a given engagement includes.
How I help: Scope documents and past proposals indexed, with budget qualified before a call.
SaaS & software
Typically needs: Setup, integration and troubleshooting questions that are in the docs.
How I help: Documentation indexed, with a ticket raised when retrieval finds nothing.
How I work
How a chatbot project actually runs.
Six steps, with a real decision point at the end of the first.
Discovery
I read your last few hundred real enquiries and work out what proportion a grounded bot could answer. We look at what content exists to retrieve from, and what does not.
Planning
The content set is decided: which pages, documents and policies are in scope, who owns each, and how often they change. Escalation rules are written down.
Design
The conversation is designed: opening message, tone, what it asks for, what it never attempts, and exactly how it says it does not know.
Development
The retrieval pipeline and the integration are built: documents chunked and indexed, the model wired up, the widget or WhatsApp number connected, logging in place.
Testing & SEO
It is tested against your real past questions, including the awkward ones, and tuned where retrieval misses. Cost per conversation is measured.
Launch & support
It goes live with monitoring, and I stay on the logs for the first weeks while the real questions arrive.
Technology
What it is built with.
Chosen per project, and explained before anything is committed to.
The AI layer
Large language modelsCommercial APIs or self-hosted, chosen on cost, latency and where the data may sit.
Retrieval (RAG)Your documents chunked, embedded and searched so answers come from them.
Vector storageA searchable index of your content, rebuilt when the source changes.
GuardrailsRules for what it must never attempt, and how it declines.
Where it runs
Website widgetA chat panel on your own site, styled to match it.
WhatsApp Business APIThe same brain on the channel most Indian customers prefer.
WordPress and WooCommerceOrder, product and policy data read from the site you already have.
Your inbox or CRMQualified conversations delivered where your team already works.
What you keep
Conversation logsEvery exchange, yours to read and export.
The content indexRebuildable from your own documents at any time.
The integration codeHanded over, documented, and not locked to me.
Cost telemetryPer-conversation cost visible rather than inferred from a bill.
Features
What gets built into every chatbot.
The parts that are not optional.
Grounded answersRetrieved from your content, not recalled from training data.
Honest refusalsA clear “I do not have that” rather than a plausible invention.
Human handoverEscalation by rule, with the conversation attached.
Conversation loggingEvery exchange stored and readable by you.
Lead captureThe fields you actually need, asked conversationally.
Source visibilityWhich document an answer came from, so you can check it.
Cost monitoringSpend per conversation, visible from day one.
ReindexingChange a document, rebuild the index, and the bot is current.
Fallback behaviourWhat happens when the model or the API is down, decided in advance.
SEO & performance
Where a chatbot helps your site, and where it does not.
A chatbot is not an SEO tactic. It sits behind a click, its answers are not indexable, and a widget that blocks content or shifts layout will cost you more in Core Web Vitals than it ever returns in engagement.
What it does give you is the best content brief available: a log of what real people asked and what your site failed to answer. Turning the top twenty of those into proper pages is the part that earns search traffic – the bot just found them for you.
- Widget loaded after content, so it cannot delay first paint
- No layout shift when the launcher appears
- Answers also published as real pages where they deserve to be
- Unanswered questions reviewed monthly as a content backlog
- No interstitial that covers content on mobile
- Script deferred and conditionally loaded
- Accessible from the keyboard and to screen readers
- Nothing the bot says is relied on for indexing
AI search
On AI answer engines, plainly.
No one can promise that ChatGPT, Gemini, Perplexity or Google’s AI Overviews will cite your site, and a chatbot on your pages has no bearing on whether they do. What is within your control is whether your content is clear, structured and attributable – which helps a retrieval system of any kind, including your own.
- Content written to answer one question per page
- Structured data that matches what the page actually says
- Facts stated once, in a place that can be cited
- The same content serving your bot and any external crawler
- Logged questions turned into real published answers
- No claim made about inclusion in any AI product
Why me
Why bring this to me.
Four things I will do that a chatbot vendor will not.
I will tell you not to build one
If your last two hundred enquiries are all different, retrieval has nothing to stand on and a chatbot will frustrate people. That conversation happens before the quote, not after the invoice.
I build the website underneath too
Twelve years of WordPress and WooCommerce means the bot can read real product, order and policy data rather than a copy that drifts out of date.
The running cost is modelled up front
Per-conversation pricing is where these projects surprise people. You see the arithmetic against your real volume before you commit to anything.
No lock-in
The index, the logs and the integration code are yours. If you want to move it in-house or to someone else, nothing here is designed to stop you.
FAQ
Questions I get asked about chatbots.
A general model asked about your business will, every time. A retrieval-grounded one answers only from passages found in your own documents, and says it does not know when nothing matches. That behaviour is the first thing I build and the first thing we test.
It depends on message volume and which model is used, and the honest answer is that I cannot quote it without your numbers. I model it against your actual enquiry volume before you commit, and the per-conversation cost is visible in the dashboard from day one.
Yes, through the WhatsApp Business API – which has its own approval process, template message rules and per-conversation pricing set by Meta. Those are real constraints and I will walk you through them before we plan around that channel.
No, and I would be wary of anyone who says otherwise. It removes the repeated questions that never needed a person and hands over everything else with context. Teams end up doing fewer, better conversations.
Usually yes, and mixed or transliterated messages are handled far better by retrieval than by a scripted bot. But this is worth testing against your customers’ real messages rather than taking on trust, and that test is part of the build.
That depends on which model you choose, and it is a decision we make together rather than one I make quietly. Self-hosted options exist where data must not leave your control, at higher cost. I will lay out the trade-off.
A fallback decided in advance: usually a message saying the assistant is unavailable and a direct route to a person. A chatbot that fails silently is worse than no chatbot.
A focused build is typically a few weeks, and the longest part is rarely the code – it is getting the content in order. If your policies live in three different versions in three different places, that comes first.
Often, yes. The usual diagnosis is that it is answering from a general model with no retrieval under it. Putting your own content underneath it is a smaller job than starting again.
Yes. Most clients want the index kept current and a monthly look at what the bot could not answer, which is also the most useful content brief you will get. It is quoted as a retainer, not bundled in silently.