- Can AI agents actually replace my support team?
No, and I'd be wary of anyone who says yes. What an agent does well is the first ninety seconds: reading the message, working out what it's about, pulling the right policy or order detail, and drafting a reply. What it does badly is judgement, anger, and the strange edge cases that make a customer loyal or gone.
So I build it as a colleague, not a replacement. It handles the routine, hands over the rest with the context attached, and a person signs off on anything that matters. Start with one queue, test it against real past tickets, then widen.
- We want to add AI to our app. Where do we start?
Start with the job, not the model. Write down one thing a user does today that is slow, boring or error-prone, and what 'done well' would look like. That sentence is your spec.
Then collect twenty real examples of it. If you can't, the feature isn't ready. If you can, we test a plain prompt against those twenty before writing any product code. Most of the time the answer to 'where do we start' is smaller and cheaper than the meeting where it was asked.
- Will my data be safe if I add an AI feature?
'Safe enough' is a design decision, not a promise anyone can make in one line. The questions I ask first: what data actually needs to reach the model? Can it be masked or left out? Who is the vendor, where do they process it, how long do they keep it, and do their terms let them train on it?
Often the answer is to send less, log less, and keep the sensitive parts inside your own system. I'll tell you plainly where the risk sits before we build, not after.
- How much does a website or an app cost?
Websites and apps usually start around ₹50,000 and go up with what they have to do: accounts, payments, integrations, editorial workflows, scale. Content and design work runs from about ₹1,000 for small pieces to ₹1.5 lakh and more for larger jobs.
What moves the price is rarely the number of pages. It's how many decisions are still undecided when we begin. Tell me what you're trying to do and I'll give you an honest range, including the version that costs less.
- My news portal is slow and my rankings dropped. What do you check first?
Three things, in this order. One: what a real reader on a mid-range phone actually experiences. Heavy scripts and ad slots do most of the damage. Two: whether the site structure still makes sense to a search engine: categories, duplicate URLs, old redirects, pages that shouldn't be indexed. Three: whether your editors can publish without fighting the CMS, because a painful workflow quietly kills freshness.
I've worked on news portals and application-based sites and improved them while they stayed live. That's the skill: changing the engine without stopping the train.
- Should I build a mobile app or just a website?
A website first, unless you can name what only an app can do: notifications people actually want, offline use, camera or location that matters, or something people open every day. If it's none of those, a good mobile-friendly site gets you further for a fraction of the cost.
If it is an app, build the smallest useful version, put it in front of real users, and let their behaviour write the second version.
- What happens after launch?
That's where most of the work is, and where most agencies disappear. Sites and apps need updates, security patches, small changes, and someone who remembers why things were built the way they were.
Several of my clients have stayed for over a decade, renewing every year. I've never had to chase one. That isn't a sales trick. It's what happens when you keep showing up after the invoice.
- Is SEO dead now that AI answers everything?
SEO isn't dead. Lazy SEO is. AI search still reads the web, and it favours pages that answer a real question clearly, come from someone who evidently knows the subject, and are easy to read and cite.
So the work shifts: fewer keyword-stuffed pages, more useful ones; clean structure and schema; and being present where your buyers actually ask questions. I'd rather write one page that deserves to be quoted than ten that merely rank.
- Which social platform should my business be on?
Where your buyers already pay attention, not where you enjoy scrolling. Selling to businesses or institutions? Usually LinkedIn, plus a website that backs up what you post. Selling to consumers? Instagram or YouTube, depending on whether your product is best shown or explained.
Pick one, do it properly for ninety days, and look at the conversations it started, not the likes. Then decide about the second. Social media management from me starts around ₹45,000 a month.
- I get leads but they never turn into customers. Why?
Usually one of three things: the wrong people are arriving, the page they land on promises something different from what you sell, or nobody follows up while the interest is warm.
Before you spend another rupee on ads, I trace one lead from first click to first conversation and see where it goes quiet. It's rarely a mystery once you look, and fixing it is cheaper than buying more of the same traffic.
- We've been in business for decades but look outdated. Do we need a rebrand?
Maybe not a rebrand. Maybe a translation. Thirty years of trust is an asset. The job is to make it visible to someone meeting you online for the first time: clear words about what you do, a face and a name, proof that you're still open and still good.
I'd start with what you say and how it looks on a phone. A new logo comes last, if at all. I come from a family business in Daryaganj that has been running for decades, so I take this one personally.
- How do I build a personal brand as a consultant or founder?
Say one true, specific thing, repeatedly, to the right people. Not 'thought leadership': a point of view someone could disagree with.
Practically: one clear sentence about who you help and how; a profile that says it in five seconds; a handful of pieces of writing that show how you think; and steady presence in one place. Consistency beats cleverness. People who hire consultants mostly want proof that you've thought about their problem before they arrived.
- Do you work with small organisations, NGOs and campaigns?
Yes. A lot of my work has been with policy research institutions, civil society networks and national campaigns, as well as businesses. Different budgets, same standard: build it properly, make it easy for their team to run, and don't leave them stranded after launch.
If your budget is tight, say so at the start. There's usually a smaller first step that still works.
- How do you use AI in your own work?
As a fast, tireless assistant that I supervise: first drafts, research, code scaffolding, testing ideas quickly. Not as a replacement for thinking, and never for anything I wouldn't put my name to after reading it myself.
This page is a small example. The curated answers are mine; the live ones come from an AI I've told to be honest, brief, and clear about what it doesn't know.
- What does working with you actually look like?
We talk first. No forms, no deck. I'd like you to tell me something real in the first five minutes: what isn't working, what you're afraid of, what you'd like to be true a year from now.
Then I write back with what I think, including what I wouldn't do. If we both like it, we start small and prove it works before we scale it. I'm not the cheapest and not the biggest. I'm the one who stays.