Illustration: a street of homes and small shops, drawn at your local time of day. Lights come on in some windows. Three of them link to the moments below.
You know a problem worth solving.
Maybe you live with it every day. Maybe you've watched it wear down people you care about. Sevarth starts there, with the people closest to a problem, and works with them toward a response that helps and can keep going.
This is how we intend to work. We don't yet have results to show that it works.
Most problems worth solving look ordinary.
They show up as a lost hour, a confusing letter, an evening that slipped away. Here are six moments from one ordinary day that many people will recognize. We chose them because we have lived versions of them ourselves.
Illustrations Written to help imagine a better day. They are not products Sevarth has built, and the people are not real clients.
Six moments from one ordinary day. Move the time on the street above, or choose a moment here.
6:45 am
The morning feed
As it is
The alarm stops and the phone is already in her hand. Forty headlines in ten minutes: a celebrity feud, a war she can do nothing about, a quiz about her personality. Nothing about the water notice for her street, the change to her son's school bus, or her own field. She gets up feeling worse, and knowing less about the things that touch her life.
As it could be
Five minutes, in her own language. The water notice for her street, and what to do. The school bus change. One development in her field, explained. One major story from the world, with what is confirmed and what isn't. And one piece she would never have picked: a farmer two continents away solving a problem a lot like hers. Each item says why it is there, and she can ask for less of anything.
When it stays unsolved
Across 48 countries and nearly 100,000 people surveyed, 40% say they sometimes or often avoid the news, up from 29% in 2017, and only 40% trust most news most of the time. Among people who avoid it, nearly one in five say it doesn't feel relevant to their lives. When people tune out, they also miss the news that would have helped them.
What it would take: news chosen for what touches your life and what widens it, with a reason shown for every item and the controls in your hands.
Who might pay: readers directly, or local and public-service newsrooms that want to reach the people they are losing. It must not be paid for with attention. Untested.
7:10 am
Before the shutters go up
As it is
Forty-one messages across four apps. A supplier's new prices are buried in a group chat. A rule about food labels changed last week, in a document nobody has had time to read. A customer waits while the owner scrolls. Her daughter's concert is at four. She probably won't make it.
As it could be
One short note each morning. Three things changed and one needs her decision, and each line shows where it came from. She deals with it in a minute, serves the customer, and closes at half past three.
When it stays unsolved
Her morning repeats in small businesses everywhere. They make up about 90% of the world's firms and more than half of its jobs. In the United States alone, federal paperwork takes a typical small business more than 379 hours a year, about ten working weeks. That time comes out of customers, family and the chance to grow.
Sources: World Bank, SME finance; American Action Forum, 2017
What it would take: trusted sources gathered in one place, summaries that show their working, and decisions left with the owner.
Who might pay: a trade association or supplier network, if missed changes cost its members money. Untested.
1:05 pm
Lunch at the desk
As it is
Lunch is whatever is closest, eaten over the keyboard between two calls. For years he has been told to eat better, but nobody has told him what that means for his body: how much, of what, given his blood sugar, his long sitting days and the food he grew up loving. Every app he tried wanted him to log every bite. He stopped after a week.
As it could be
Nothing to log. With his permission, his usual lunch order now leans toward what his body needs today, in a portion that fits his day, and it still tastes like home. Once a month he sees a simple trend and one small change to try. He didn't have to change everything at once. He started by eating what he needs, and about as much as he needs.
When it stays unsolved
In 2024, one in seven people in the world, about 1.1 billion, were living with obesity, including 170 million children and young people aged 5 to 19. Many of the same countries still face undernutrition. Most advice is the same for everyone, while what a body needs is personal.
Source: WHO, Obesity and overweight
What it would take: guidance personal to each body and food culture, built with clinicians and nutrition scientists, that adjusts quietly instead of asking people to track everything.
Who might pay: employers, insurers or health systems that already carry the cost of diet-related illness. Medical decisions stay with clinicians. Untested.
6:40 pm
The bus route
As it is
A flyer says the route ends in March. The neighborhood chat says the plan was cancelled. The city's proposal runs to sixty pages. The new stop would be two streets away, at a curb with no ramp. He has something important to say. By Thursday's meeting he still isn't sure what is true, so he stays home.
As it could be
One page sets out what is decided, what is proposed and what is still open, each with its source. There is a clear place to send evidence, so he sends a photo of the curb. On Thursday he goes to the meeting knowing what he wants to ask.
When it stays unsolved
About 1.3 billion people, one in six, live with a significant disability. For them, transport that is inaccessible or unaffordable is 15 times more of a barrier than for others. Only about half of the world's city dwellers have convenient access to public transport at all. When the people a decision affects can't follow it, it gets made without them, and trust wears away: 44% of people across OECD countries have low or no trust in their national government.
Sources: WHO, Disability and health; UN SDG Report 2023; OECD Trust Survey, 2024
What it would take: plain-language comparisons with sources, and a public process that can take in evidence from the people a decision affects.
Who might pay: a city service that needs better input before it decides. Untested.
8:15 pm
After dinner
As it is
The dishes are nearly done. On the sofa, one video ends and the next starts on its own. "How was school?" "Fine." Nobody chose this evening. It just happened again.
As it could be
At dinner she asked why bread rises. Now a jar of yeast and flour sits on the table, and the tablet is asking her questions back. Warm water or cold? What do you think will happen? Her dad leans in. He doesn't know either.
When it stays unsolved
Across OECD countries, 45% of 15-year-olds feel nervous or anxious when their phone isn't near them, and 65% are distracted by devices in at least some maths lessons. Students who spend up to an hour a day on devices for fun score 49 points higher in maths than those who spend five to seven hours, even after accounting for family background. That is a link, not proof of cause, but the gap is large, and the evenings add up to a childhood.
Source: OECD, PISA 2022 results
What it would take: tools built around curiosity and judgment rather than time spent watching, shaped by the parents and teachers who know the children.
Who might pay: schools, libraries or families. It must not depend on advertising or time on screen. Untested.
10:30 pm
The kitchen-table accounts
As it is
The rent went up in spring. Two subscriptions renew this week that nobody remembers choosing. The school fees are due in March, and the money set aside for them has slowly gone into ordinary months. They find out when a payment bounces. The argument that follows isn't really about money. It's about not having seen it coming.
As it could be
Weeks ago, an assistant that works only for them noticed the gap opening: at this rate, they would be short before the fees were due. It had handled the small things quietly. Now it spoke up once, with three options worked out for their situation, each with its trade-offs: cancel two subscriptions, move the fee money into its own pot, or take one extra weekend shift for two months. They chose. It did the rest.
When it stays unsolved
In the United States, 37% of adults could not cover a $400 emergency with cash or its equivalent. In low- and middle-income economies, only about 55% of adults could come up with extra money in an emergency. Most households find out about a shortfall when it arrives, not while it could still be avoided.
Sources: Federal Reserve, household well-being in 2024; World Bank, Global Findex 2025
What it would take: an assistant that works for the household, not for a lender or a seller. It watches quietly, speaks up early, explains the options in plain words, and leaves the decision with the people it serves.
Who might pay: households directly, or employers and credit unions offering it as a benefit. It must never earn money by selling them financial products. Untested.
We are in these windows too. The problems we choose to solve now shape the ordinary days our future selves will live, and the ones the next generation inherits.
We can build almost anything now. Choosing what deserves to be built is harder.
Software that once took a team a year can now take a few people a few weeks. That is extraordinary. But capability doesn't tell us what deserves our effort, and attention flows easily toward what is new and impressive. Problems that shape ordinary days stay unsolved because they are unglamorous, or because the people who understand them are not in the room.
An observation from two decades of working in technology, not a measured figure.
What problem, if solved, would make life meaningfully better?
What we believe technology owes people.
Give more than it takes.
Technology should leave people with more than it takes from them, in money, attention and time. Much of today's technology is paid for with attention, and products are tuned to hold it. There are better ways to create value, and better ways to be paid for it.
Our founding conviction, and the test we apply to every idea.
Stay quiet until it matters.
The best help works in the background, like good plumbing. It handles what it can and speaks up only when something needs a person's attention or decision, early enough to act on it.
This idea has a name: calm technology, described by Mark Weiser and John Seely Brown at Xerox PARC in 1995. Read the paper.
Offer what people need, not only what they ask for.
Sometimes the most valuable thing is what you would never think to ask for, because you have never been exposed to it. Technology should widen people's world, not narrow it. So it must always say why it is showing something, make it easy to say no, and never let a seller decide.
This kind of suggestion can also be used to manipulate. That is why the conditions are part of the principle, not an afterthought.
Two words that share a letter.
Seva and arth share the letter a. Joined at that letter, they make Sevarth.
Seva is usually translated from Sanskrit and Hindi as service. Arth (artha in Sanskrit) carries meaning and purpose, and also the material means to live well. Joined at the letter they share, they make Sevarth. That shared a is our mark, because the work belongs where service and lasting value meet.
The meanings follow standard references on Sanskrit terms. The idea that the work lives where they meet is our own reading, not an etymology.
Seva
Service that starts with real needs. People who live with a problem help define it, test responses and decide what counts as better. Nobody arrives to rescue anyone. Everyone involved has a stake.
Arth
Purpose with the means to last. A response that helps but can't pay its way fades when a grant ends or a volunteer burns out. So from the start we ask who benefits, who pays, and why.
Sevarth was founded by Devesh Ginodia after more than twenty years in technology, data, AI and enterprise transformation. That work showed how much technology can do. Sevarth puts the same experience to a different question: what should it be used for?
The outlook behind it draws on Indian philosophy about duty, action and responsibility. In practice it comes down to principles most people share. Do the work in front of you with care. Own its consequences. Leave things better for the people who come after you, including the person you will be.
The founder's convictions. They guide our choices; they are not claims anyone has to accept.
How an idea earns its next step.
Every problem we take on moves through five stages. Each stage has to earn the next, and some ideas should stop. When the evidence says go back a step, we go back.
This is our intended process. If it changes, this page will say so.
Three questions travel with every stage
- Value
- Is this worth doing for the people affected, and what evidence would show it?
- Trust
- Are we keeping our commitments, and respecting people's dignity and choices?
- Transparency
- Can anyone see our evidence, assumptions and incentives, and question them?
Transparency starts here. labels every claim on this page.
-
A problem worth solving
Who lives with it, how often, and what it costs them, described in their own words.
Moves on when solving it would make ordinary days meaningfully better.
-
Understanding and evidence
What is already known, what has been tried, and what failed and why.
Moves on when we can show what we know and say plainly what we don't.
-
A response you can test
The smallest version that could help, tried with the people affected.
Moves on when it helps, by their measure, not ours.
-
A model that pays its way
Who benefits, who pays, why they would pay, and what it costs to deliver.
Moves on when it can last without shutting out the people it is for.
-
Growing responsibly
Reaching more people only as fast as quality and trust allow.
Keeps going while we keep our promises to the people already using it.
Good intentions don't pay for delivery.
A problem that matters doesn't automatically come with a way to fund the answer. So for every idea we ask four plain questions, early, and we write the answers down.
- Who benefits?
- Who pays, and why would they?
- What does it cost to deliver, and to keep delivering?
- How do people who can't pay still get access?
Our first model, still being tested: organizations that already carry the cost of a problem, such as a school network, a city service, an employer or a trade association, pay for focused work to understand it and test a response. What we learn goes back to the people affected and, with their consent, into the open.
Not yet validated. We are testing two things: whether organizations will pay for this, and whether the people affected find the result useful.
We think purpose and profit can pull in the same direction when incentives are clear. Where they pull apart, we would rather say so than hide it.
Where Sevarth is today.
As of
In place
- A founder with more than twenty years in technology, data, AI and enterprise transformation.
- A way of choosing and testing problems, set out on this page.
- First areas of inquiry: the everyday moments above, from news and food to household money, work and civic life.
Not yet
- Products, pilots or results.
- Partners or customers to name.
- A community. That is what the invitation below is for.
We will update this list as things change, including when something we try doesn't work.
Start with one small step.
If something here sounds like your life or your work, tell us. It takes a few minutes, and you choose how much to share.
How we handle what you share
- We only ask for what we need to understand your note and, if you want, reply to it.
- Your email is used to reply about this note. It is never added to a mailing list or shared.
- We will never quote you without asking. Nothing is pre-ticked.
These are promises Sevarth intends to keep. Hold us to them.
Thank you. Your note is ready.
Nothing has been sent. This preview of the Sevarth site has no way to deliver messages yet, and we haven't published a contact address here. Copy your note and keep it. When the site goes live it will show where to send it.