A case study and story, in the words of Gautam Shah, founder of Indigo Events and Promotions, twenty years in events.
Who I am
My name is Gautam Shah. I have spent more than twenty years in the events business. Not behind a desk writing about events, but on the floor running them. Weddings, corporate launches, conferences, award nights, festivals. I have built mandaps at 4 in the morning. I have argued with vendors who did not show up. I have stood at a venue at 2 am when the generator died and three hundred guests were about to arrive. My company, Indigo Events and Promotions, is a working agency. We do real events for real clients.
So when I talk about event software, I am not guessing. I have lived the pain.
Kunjara OS is the product I always wished existed. We built it because nobody else built it the way an actual event person needed it.
The problem I lived for twenty years
Here is the honest truth about how most event companies run, even today.
The proposal goes out as a Word file or a PDF that takes two days to make. The budget sits in an Excel sheet that breaks every time someone touches it. The client approves over email, or worse, over a phone call you cannot prove later. The guest list is another spreadsheet. On the event day itself, the whole operation runs on WhatsApp groups and shouting on walkie talkies. Nobody has one clear picture of what is happening. You only find out about a problem when it is already a fire.
And the worst part. Every single event starts from zero. You forget which vendor was reliable last time. You forget what the client liked. You forget which crew member actually performed. All that hard earned knowledge lives in your head, and your head is busy.
I always felt that an event company is sitting on a goldmine of information and using almost none of it.
Why AI, and why most event software gets it wrong
When AI became good enough to be useful, a lot of people rushed to bolt it onto everything. Most event tech I saw used AI as a gimmick. A chatbot here. A logo maker there. Nice for a demo, useless on the floor.
I had a different belief. AI should not be the product. The event is the product. AI should quietly do the boring and the heavy work so the human can do the human work, which is creativity, relationships, and judgement.
So the rule we set for Kunjara OS was simple. Every feature must produce real operational information that the planner can act on. AI earns its place only if it saves time or removes risk on a real event.
That single rule shaped everything.
What Kunjara OS actually does
Kunjara OS follows the real life of an event, from the first enquiry to the moment the last guest leaves and the report is sent.
The proposal, in minutes instead of days. A planner types a short brief. Type of event, city, guest count, budget, a few lines of vision. Our AI, which we call Atlas X, produces a complete client ready proposal. Concept, an itemised budget with correct local tax, a vendor list, a timeline, a compliance checklist, even a mood board. What used to take two days now takes a few minutes. The planner edits, puts their own logo on it, and sends it.
The Event Room. Instead of email chains, the client opens a private online room. They read the proposal, comment on it, ask for changes, and approve, all in one place, with no login. The planner finally has a clear record of who said yes and when.
Guest Layer and KunjaraEye. Every guest gets a personal QR invite and their own RSVP page. The planner watches confirmations come in live, with headcount and dietary needs rolling up automatically. On event day, any phone becomes a scanner. We call it KunjaraEye. Point it at a guest QR and it checks them in, green or red, no extra hardware. After the event, the system writes a Post Event Report, attendance, no shows, walk ins, when people actually arrived, and shares it with the client inside the same Event Room.
Event Day Live. This is the part I am most proud of, because it solves the problem I felt most painfully for twenty years. Event Day Live is a single screen for the person running the show. A live health banner, green, amber, or red. The guest flow at the door. The run of show, what is happening now and what is next. The crew, each one tapping their own status from their phone. And a live command feed that becomes the official log of the event.
At the top of that screen sits the AI Event Director. It does not show you data. It tells you what to worry about right now. In plain language. Something like, the main stage is running twelve minutes late, the catering team has not checked in, and a VIP is arriving in eighteen minutes. That is the difference between finding out about a fire and seeing the smoke before it starts.
The Vendor Network. A searchable, verified directory of event vendors. To keep it honest and free of fakes, every vendor proves they are a real business, either with a tax number or with a small one time verification fee. No fake vendors, only real ones.
The AI under the hood, for the technical reviewers
For the people judging the technology, here is what is actually happening, in plain terms.
Kunjara OS does not rely on a single model. It uses more than one AI provider and chooses the right one for the job, with automatic fallback so the planner never sees a dead end. We use Google Gemini and OpenAI together.
The proposal engine uses structured generation and strict function calling, so the AI does not return loose text, it returns clean, validated data that drops straight into a budget, a timeline, and a vendor list. The numbers add up because the system forces them to.
The vendor suggestions use grounded search, so the AI returns real vendors for a real city and event type, not invented names.
Mood boards and event visuals are produced with image generation, so the concept is not just words, the client can see it.
The AI Event Director is the piece I find most interesting. It reads the live operational state of an event, the crew statuses, the door numbers, the run of show against the real clock, and it reasons over that to surface risk. It is AI pointed at a live operation, not at a chat box. As we add incidents, alerts, and timelines, the same Director gets sharper, because it is reading more of the truth.
And the whole thing runs as a fast, modern web application, so a planner can open it on a laptop in the office and on a phone at the venue, with no installation.
What is genuinely new here
Plenty of tools do one slice. A proposal tool. An RSVP tool. A walkie talkie app. What I had never seen is one system that follows the event all the way through and connects every slice with the same thread of information.
The brief becomes the proposal. The proposal becomes the Event Room. The Event Room becomes the guest list. The guest list becomes the live door count. The door count and the crew and the run of show become the live command screen. And all of it becomes the report and, soon, the memory.
That connection is the innovation. The AI is not a feature you visit. It is the intelligence running quietly along that whole thread, helping at every step.
Where we are, honestly
I will not dress this up with big numbers, because that is not how I work. We are early. A small group of planners use Kunjara OS today, including my own agency on real events. The product is live at kunjaraos.com. It is built for India and Singapore first, and it is open to the world.
What I can tell you with full honesty is the effect on the floor. A proposal that used to eat two days now takes minutes. A client approval that used to live in a lost email now lives in one clear room. A guest check in that used to need printed lists and a person at a desk now happens with a phone. And an event day that used to run on noise and luck now has one screen telling the person in charge where the risk is.
That is not a demo. That is a working agency using its own tool because it actually helps.
What comes next
The vision is to make Kunjara OS the system an event team opens at 6 in the morning and keeps open until teardown. Slack, a flight control tower, a walkie talkie, a run sheet, and incident management, all on one screen.
We are building that out now. Incident reporting with one tap. Crew availability. A live timeline where a delay shifts everything downstream automatically. An emergency mode for the moments when nobody should be searching a WhatsApp group for the hospital number.
And the layer I believe will matter most over time is operational memory. Vendor memory, which supplier was reliable and paid on time. Client memory, what each client prefers and how they spend. Crew memory, who is good at what. So that no event ever again starts from zero. That is the part that makes the software smarter every year, and it is the part no competitor can copy, because it is built from your own events.
Why it matters
I built this as an event person, for event people. The industry runs on the backs of teams that work brutal hours and carry enormous stress, often invisibly, so that a wedding or a launch looks effortless. Good software should respect that work and lift it.
Kunjara OS is made in Bharat and built for the world. It uses AI not to replace the planner, but to give the planner what they have never had, which is calm, clarity, and a memory that lasts.
That is the story. Twenty years of standing at venues taught me exactly what was missing. AI finally made it possible to build it.
Gautam Shah
Founder, Kunjara OS
Indigo Events and Promotions
Fact sheet
- Product: Kunjara OS, an AI operating system for event planning and execution.
- Founder: Gautam Shah, twenty plus years in events, founder of Indigo Events and Promotions.
- Live at: kunjaraos.com.
- Markets: India and Singapore first, open worldwide.
- What the AI does: generates complete client ready proposals from a one line brief, suggests real vendors, creates mood boards, analyses guest lists, and runs an AI Event Director that reads a live event and tells the organiser what to worry about right now.
- AI approach: multiple models, Gemini and OpenAI, with structured generation, strict function calling, grounded search, and image generation. AI serves the workflow, it is not a gimmick.
- Core modules: AI proposals, Event Rooms, Guest Layer with QR check in (KunjaraEye), Event Day Live cockpit with the AI Event Director, Vendor Network, budgets, and compliance.
- What is new: one connected system that follows the event from the first brief to the post event report, with AI running quietly along the whole thread.
- The vision: operational memory, so no event ever starts from zero.