Service 03 — Chatbots + workflows
AI that saves hours, not AI for the sake of it.
We build three kinds of AI systems: chatbots that answer customers from your real business content (on your website and WhatsApp), automations that take over repetitive work like quotes, follow-ups and data entry, and AI features integrated into your own product or internal tools.
Every project starts from a workflow that measurably eats time, and ends with a number: hours saved per week. If we cannot find that workflow on the discovery call, we will tell you AI is not what you need yet — that honesty is cheaper for both of us.
What's included
What you get
Chatbots grounded in your content
The bot answers from your documents, price lists and policies — not from the model's imagination. When it is not confident, it says so and hands the conversation to a human.
WhatsApp & website deployment
Deployed where your customers already are: embedded on your site, connected to the WhatsApp Business API, or both — with the full conversation history visible to your team.
Workflow automation
Enquiry-to-quote, lead follow-up sequences, invoice processing, moving data between the tools you already use — automated end to end with a human checkpoint where it matters.
Custom AI integration
Document processing, search across your own data, drafting and classification inside your product — built on Claude or GPT-class models chosen per task.
Testing against reality
Before launch, the system is tested against real past conversations and documents from your business — not toy examples — and tuned until the failure modes are boring.
Cost control built in
Model usage is metered with a monthly cap you set. You always know what the system costs to run, and a busy month never produces a surprise bill.
How it works
The process
01
Find the workflow
On the discovery call we look for the task that eats hours every week — repeated enquiries, manual quotes, copy-paste between tools. That one workflow becomes the project.
02
Prototype on your real data
Within one to two weeks you interact with a working prototype fed by your actual content and examples — the fastest honest way to see whether the idea holds up.
03
Build & harden
We productionise the prototype: edge cases, human handoff, logging, cost caps and the boring engineering that separates a demo from a system you can trust with customers.
04
Launch with a safety net
The system goes live with human oversight — your team sees what it is doing and can step in. Confidence is earned in production, not assumed.
05
Measure & expand
We review the numbers after the first month: hours saved, conversations resolved, errors caught. Then you decide — from evidence — whether to automate the next workflow.
Why it pays off
What this means for you
Answers at 2am, in any timezone
A grounded chatbot handles the same twenty questions your team answers every week — instantly, around the clock — and escalates the conversations that actually need a human.
Payroll-sized savings
An automation that saves ten hours a week pays for itself in months. We put the number on paper before you commit, and measure it after launch.
No vendor lock-in
Built so the underlying AI model can be swapped as better or cheaper ones ship. You benefit from the fastest-moving market in software instead of being trapped by it.
Your data stays yours
Commercial API terms that do not train on your data, documents kept in your own storage, and masking for sensitive fields — put in writing in our agreement.
Typical investment
₹50,000 – ₹5,00,000+
≈ $600 – $6,000+
A content-grounded chatbot typically runs ₹50,000–₹1,50,000; workflow automation ₹75,000–₹2,50,000. Running costs are billed at cost with a monthly cap. Details in the pricing guide.
Good questions
FAQ
Will an AI chatbot actually answer correctly about my business?
Yes, if it is built properly. We ground the bot in your real content — your site, documents, price lists, policies — so it answers from your material rather than making things up. It is tested against real customer questions before launch, and when it is not confident, it says so and hands over to a human on WhatsApp or email instead of guessing.
Which AI models do you use?
We pick per task — typically Claude or GPT-class models, balancing quality against running cost. Your integration is built so the underlying model can be swapped as better or cheaper options appear, without rebuilding anything. You are never locked to one vendor's pricing.
What does an automation project look like at the start?
We start with one workflow, not a grand AI strategy. On the discovery call we find a task that eats hours every week — answering the same enquiries, preparing quotes, copying data between tools — and scope an automation around it. You see the time saved within weeks, then decide whether to automate the next thing.
Is my business data safe? Will it be used to train AI models?
We use commercial AI APIs whose business terms do not train on your data — this is standard for the paid tiers of Anthropic and OpenAI, and we put it in writing in our agreement. Your documents stay in your own storage, we access only what the automation needs, and we can add masking for sensitive fields like customer contact details.
What are the ongoing costs of an AI system?
Two parts: model API usage (you pay per conversation or document processed) and hosting. For a typical small-business chatbot that is roughly ₹500 – ₹5,000 per month depending on volume, billed at cost. We set a monthly cap so a busy month never produces a surprise bill.
Have a project in mind?
A discovery call is 30 minutes, free, and useful either way — you leave with a clear scope and an honest number, whether or not we work together.