The Runway Was Short. Simtel.ai Kept Building.
From AI Possibilities to Real Business Impact.
August 2023 — It Started With a Short Runway
Three years ago, Simtel.ai began with an idea, a belief in artificial intelligence and a willingness to build. The company was bootstrapped, the runway was short, and there was little certainty about which ideas would survive long enough to become products. There was no perfect map to follow. So, Simtel.ai did what many young companies do when time and resources are limited: it kept building.
The early months were spent experimenting, learning and speaking to businesses with very different problems. AI was moving quickly, but knowing that a technology was powerful was not the same as knowing where it could be useful. Slowly, the company began to understand that the technology itself could not be the answer. The real value would come from solving meaningful problems.
Late 2023 — More Room to Build
Support arrived through Microsoft for Startups, giving Simtel.ai access to the infrastructure needed to continue experimenting. Over the course of the journey, the company received more than $60,000 in credits, including a $50,000 allocation on February 4, 2026. For a bootstrapped company building with AI, access to computing infrastructure meant more than a milestone. It created room to test, learn and continue building.
The support came at a time when every resource mattered. The company was still exploring different directions and trying to understand where its capabilities could create the most value. There was no finished product portfolio yet, only a growing collection of experiments and lessons. The runway was still short, but it had become a little longer.
2024 — From AI Ecosystems to Real Problems
In 2024, Simtel.ai joined NVIDIA Inception, becoming part of a wider ecosystem of startups building with artificial intelligence. It was an important moment in the company’s early journey, but ecosystem support could only take the company so far. The harder work was still ahead: finding real problems and building something useful around them. Technology had to meet the market eventually.
One of those problems came from the world of D2C commerce. Customers had more choices, brands were competing for attention, and finding the right product was becoming almost as difficult as buying it. Simtel.ai began to see commerce as more than a transaction problem. It was increasingly a problem of discovery, context and intelligence.
From D2C to AI-Commerce
That thinking led to Gallyria and a deeper exploration of AI-commerce. Gallyria was built around the idea that product discovery could become more intelligent, contextual and conversational. The company’s work with ONDC added another dimension, connecting AI-driven commerce with India’s open digital commerce ecosystem. The question was no longer simply how to sell products online, but how to help people discover the right ones.
By late 2024, Gallyria had become one of the visible outcomes of that work. But building products and building a business were not the same thing. For much of Simtel.ai’s early journey, the focus had been on technology, experimentation and solving problems. The commercial engine had yet to fully catch up.
Early 2025 — The Commercial Engine Arrives
That began to change in early 2025, when a co-founder joined Simtel.ai and brought a stronger focus on go-to-market and sales. The conversation inside the company began to shift from what could be built to who needed it, how it should be positioned and how it could reach them. The builders were still building, but the company was beginning to learn how to sell.
For a bootstrapped company, this was a necessary transition. Products could not remain experiments forever, and technology alone could not create a sustainable business. Client conversations, market positioning and revenue began to take on greater importance. The company was beginning to connect what it built with the people who needed it.
AxonOS — Building the Intelligence Layer
As Simtel.ai began working across products and use cases, it was also thinking about the infrastructure beneath them. In 2025, AxonOS emerged from that effort, with a framework for bringing intelligence, workflows and execution together. It reflected a growing realization that AI applications needed more than models and prompts. They needed structure, orchestration and a way to connect reasoning with action.
That idea is now moving into its next phase with AxonOS v2. The new direction is centered on a harness-driven orchestration platform, designed to bring greater control and structure to intelligent and agentic systems. What began as experimentation is gradually becoming a more deliberate architecture. The lessons from building across different products are now shaping the platform’s next version.
RiskOps — Looking at What Comes Next
Another direction began to take shape around a different problem: risk. Risknob.com, the upcoming RiskOps platform, is being built around signals, uncertainty and the risks organizations need to understand before they become larger problems. It is still preparing for launch, making it less a milestone from the past than a bet on the future.
Even after three years, Simtel.ai is still building new things. But the experiments are no longer starting from the same blank page as before. They are increasingly informed by earlier products, client conversations and a clearer understanding of where AI can be useful. The exploration continues, but with more context behind it.
The Latest Chapters
The journey has now moved into more specific domains. ConsultLegal.in grew from the experience of working with a legal firm and the realization that legal intelligence depends heavily on documents, context, knowledge and professional workflows. It is another example of moving away from generic AI and toward systems designed around the needs of a particular industry.
ConsultLegal.in brings together AI with legal knowledge and workflows, while regional language capabilities, OCR, translation and embeddings help strengthen its contextual intelligence. Sarvam is part of that broader technology effort, supporting capabilities required to work across languages and information formats. The goal is not simply to generate answers, but to work with information in a way that is more useful within its context.
Then came an unexpected direction: coffee. Through AEC-CCRI-CED and the Coffee Board ecosystem, Simtel.ai began working around branding and coffee exports for tribal estates. The challenge extended beyond technology itself, involving visibility, market access, storytelling and connections to potential buyers.
That work is evolving into Coffex.org, focused on intelligence for the coffee ecosystem. The broader vision includes market intelligence, buyer-seller discovery, sourcing, traceability, branding and export opportunities. It is a different industry, but the approach remains familiar: understand the context, identify the problem and apply technology where it can create value.
What Connected All of It?
Looking back, the journey does not resemble a conventional corporate plan. It moves from a bootstrapped AI startup to NVIDIA Inception, from D2C work to AI-commerce, Gallyria and ONDC, and then toward AxonOS, RiskOps and Risknob.com. The latest chapters, ConsultLegal.in and Coffex.org, take those lessons into highly specific industries with very different needs.
On paper, these directions can appear unrelated. But from inside the company, each step followed the same instinct:
Find a real problem. Understand the context. Build something useful. Then learn from what happens next.
The first three years were not simply about launching products. They were about discovering where AI could actually work, where it needed more structure and context, and where technology could move beyond demonstration into something useful. Every client, experiment and platform added another piece to that understanding.
From Building to Business
For much of those first three years, survival was part of the work. Simtel.ai was bootstrapped, and the runway was never long enough to treat experimentation as an abstract exercise. Products had to become useful, and useful products eventually had to become sustainable businesses. Building was necessary, but building alone was never going to be enough.
The stronger focus on GTM and sales in 2025 changed the nature of the journey. Simtel.ai was no longer only accumulating technology and ideas; it was beginning to connect what it had built with customers, markets and revenue. The shift from building to selling did not happen overnight, but it gave the company a commercial engine to match the technology that had been taking shape.
2026 — The Next Phase
By 2026, the pieces are beginning to converge. Gallyria continues the company’s work in AI-commerce and ONDC, while AxonOS is moving toward version 2 and Risknob.com is preparing for launch. At the same time, ConsultLegal.in and Coffex.org represent the newest applications of what Simtel.ai has learned about contextual intelligence and domain-specific AI.
This year also brings a different financial ambition. After three years of building with a short runway, Simtel.ai is focused on becoming self-sufficient and profitable. The goal is no longer simply to extend the runway long enough to reach the next milestone. It is to turn the products, client experience and lessons accumulated over three years into sustainable businesses.
That may be the real story of this anniversary. The first three years were spent exploring the possibilities of AI, testing where it could work and learning what businesses actually needed. Some ideas became products, some products evolved, and some opportunities arrived from places no one could have predicted. But every step added to the company’s understanding of what works.
The Next Chapter
Three years ago, Simtel.ai was asking a simple question:
What can we build with AI?
Today, the question is different. After working across commerce, ONDC, orchestration, RiskOps, legal intelligence and coffee, the focus is increasingly on applying what has been learned to real problems, real customers and sustainable businesses.
The runway was short. The learning was long. And the years of building have produced something more valuable than a list of milestones: a clearer understanding of what works.
By,
Ravi Kaushik & Sunil Kumar Perumal