FOUNDER'S OFFICE · INSPIRELABS · HYDERABAD

Three seats. A line, not a lane.

We are not hiring you to assist. We are hiring you to run something that does not exist yet. No resume required.

Seats:

3 Experience:

2–5 years (1–2 considered on evidence) Location:

Hyderabad, in office Apply with:

Proof of work, not a CV

Most Founder’s Office jobs are a lie.

You get the title, a seat near the founder, and eighteen months of building decks about work other people are doing. Then you leave with a good LinkedIn headline and nothing you can point at.

This is the opposite of that, and it will suit far fewer people. Read the whole thing before you decide.

Who we actually are

Most people know one of our products. There are six.

We are InspireLabs. We started in 2013 with no funding and we still have none. Nothing here was bought. All of it was built, by a small team, out of cashflow.

GrabOn was product one. Six million shoppers a month, 3,800 plus brands, over ₹2,700 Cr unlocked for people, running in India, the US and the UAE. It is a real business and it pays for everything below it.

Everything below it is the interesting part.

- GrabCash.

Affiliate income without the influencer industry. Anyone with a network shares a link and gets paid to their bank. Server-side attribution.
- GrabShare.

Creator commerce. Revenue-share storefronts for people who have an audience and no way to make money from it that does not feel grubby.
- GrabCredit.

Being built right now. One of the seats below could end up owning a piece of it.
- And one more.

Founder-led build, deliberately unannounced. You will hear about it in the second conversation, not in a job ad.

The part that makes this unusual: every surface in the portfolio sits on a loop that runs itself. Research, decide, execute, learn. A fleet of agents does the work and humans stay in the loop where judgment actually matters. Not chatbots. Workers.

What the job actually is

You get a line, not a lane.

You will be handed one initiative. Not a workstream inside somebody else’s initiative. A thing, with a name, that either works or does not, and it will be obvious which.

You run it the way a founder would run it, except you do not have to raise money, find a co-founder, or build the platform underneath. You inherit twenty-five million monthly users, a merchant network, an affiliate stack, an agent fleet and thirteen years of distribution. Then you go and make something out of it.

In twelve months you are either running that line with a team under you, or we killed it and you are on the next one.

Your first 90 days

- Weeks 1–2.

You take your initiative and get it in front of reality fast. Real users, real merchants, real creators, real money. Not a market study. If your first artifact is a document, you have started wrong.
- Weeks 3–6.

You build the loop. The agents that do the work, the number that tells you if it is working, and the smallest version of the product a stranger will actually use. You will write code. Everyone here does now.
- Weeks 7–12.

Something is live. You defend the number weekly, in front of everyone, including when it is bad. Then you either scale it or you kill it out loud and take the next one.

You will sit close enough to me that there is no translation layer. What you decide on Tuesday can be live on Thursday. That is the upside, and it also means there is nobody to hide behind,



which is the part people underestimate.

Three seats. Same wiring, pointed in different directions. Seat 1 · Runs a line

You take one of the new products from a sentence to something with users and revenue. You hold three or four threads that do not obviously connect and you see the connection first. You size a bet, you take it, and you cut it fast when it is wrong.

You are not doing this from a document. Your hands are in the product. If you need six weeks of analysis before you will commit to a direction, this seat will make you miserable and you will make everyone around you slow.

Ends up owning:

a product line, P&L; attached.

Seat 2 · Builds the machine

You look at anything done manually more than twice and you cannot leave it alone. You take a job that currently costs six people a week and turn it into a fleet of agents that does it overnight, correctly, with the humans moved to the two decisions that actually needed a human.

You get handed something messy involving four people, none of whom report to you, and a date. You make it happen anyway without being unpleasant about it, and those four people want to work with you again.

Ends up owning:

the agent layer across products.

Seat 3 · Brings the demand

New products die of no distribution, not bad code. GrabCash and GrabShare need brands with budgets on one side and creators with real audiences on the other. You find the person who actually has authority, you get through, and you make them want it.

You have done this with no brand behind you, which is the only version that proves anything. You will also be the one who has to kill a deal that looks good and is not.

Ends up owning:

supply and demand for the new lines.

Apply to one. If two genuinely fit, say so and say which is stronger. Do not say all three. We will not believe you.

How we will know it is you

We are looking for a particular kind of wiring, not a particular CV. Some of these will feel oddly personal. That is deliberate.

- When you do not know a number, your first move is to build a small thing that tells you. Not to schedule a meeting about it.
- You get restless when things go quiet. Stability makes you itchy and you go looking for the next problem before anyone assigns it.
- You have shipped something end to end with nobody checking on you. It may have been small. It may have failed. It was yours.
- You do not say a number out loud until you have checked it twice. When you are wrong you say “I was wrong” in four words and move on.
- You have taken charge in a room where you had no authority to, and it was not weird.

What will actually be hard here

- Bootstrapped.

There is no burn to hide inside. Every rupee your initiative spends was earned by another product downstairs, and everyone knows it.
- Recent things get killed.

Some of what you build will not survive contact with the market. You have to be able to be the person whose thing got killed, say so publicly, and start the next one on Monday.
- Ambiguity is not a phase.





It does not end after onboarding. It is the medium.
- We measure weekly.

If you have a slow month, we will both know.
- I am direct.

Some people find that bracing. If you need feedback wrapped, we will both be tired by month two.

Who this is not for

We will not hire a strategist.

If your value sits in the analysis and somebody else does the doing, this is the wrong company and, honestly, the wrong decade. We are hiring operators. People who finish things.

We do not filter for colleges.

We do not read that line. Some of the best people here came from places you have never heard of, and a few came from places you have and it made no difference either way. If your strongest argument is where you studied, you will find us frustrating.

Two to five years of doing real things is roughly the range. If you are at one or two years and your work is clearly ahead of your experience, apply anyway. We will be able to tell in about ten minutes.

The application

We are not asking for a resume. You can send one. We will not open it first.

Pick one problem below. Spend up to five days. Send us what you built. These are not puzzles. They are the shape of the actual work.

Problem 1 · Build a worker, not a chatbot

Pick a repetitive commercial job a small company would currently pay a person to do all day. Qualifying brands. Finding creators with real audiences rather than bought ones. Watching competitor pricing. Turning a messy public dataset into decisions.

Build an agent, or a small fleet, that does the whole job unattended overnight and produces an output somebody could act on Monday morning without cleaning it up first.

Ship:

the code, the output, and one place the model was confidently wrong that you caught.

Problem 2 · Find the wedge

Pick one of our live products: GrabCash, GrabShare, RankDrive or Alternatives.co. Use it properly, as a real user, for a few days. Not a teardown from the outside.

Come back with the single thing you would change in your first 30 days if you owned it, and the evidence. Then the part we actually score: the three obvious things you would refuse to build, and why.

Ship:

the evidence, the one change, the three refusals.

Problem 3 · Get the yes

GrabCash and GrabShare need two things: brands with budgets, and creators with real audiences. Pick a side. Find ten genuine candidates and explain how you found them.

Then reach the actual decision maker with nothing behind you, and get one substantive reply. Not a polite brush-off. A reply with something in it.

Ship:

the research, the exact message you sent, and the reply. We verify it.

How to send it

Email [email protected] with the subject line:

SEAT [1/2/3] — PROBLEM [0/1/2/3] — YOUR NAME In the body: what you built, what you got wrong, and what you would do with two more weeks. Under 200 words. Attach or link the artifact. Add a three minute video walking through it. Phone camera is fine, we are not scoring production quality.

What happens after

- We read everything inside 5 working days. We reply to everyone, including the noes.
- 45 minutes taking your work apart. Bring your reasoning, not slides.
- A paid work day on a real initiative, with the team you would be joining.
- A conversation with me, where you find out what the unannounced one is. No structure, no panel.
- Offer, or an honest no, within three weeks of your first email.

📌 FOUNDER'S OFFICE · INSPIRELABS · HYDERABAD | Three seats
🏢 GrabOn
📍 Hyderabad