Mukund or Madhav Jar

Co-founder of Emergent, an AI platform for software development.

Claims

As speaker

Mentioned in

Key positions and views

The future of software is 'agent-first,' and existing SaaS companies that fail to pivot will become obsolete.

The core technical challenge in creating autonomous AI developers is not code generation, but verification, which enables agents to operate reliably over long periods.

Foundation models are becoming a commodity; the defensible moat lies in the orchestration, infrastructure, and verification layers built on top of them.

The largest untapped market for software creation consists of non-technical individuals, and the goal is to provide a platform that abstracts away all programming complexity.

A deeply integrated, customer-centric culture, where even engineers speak to users weekly, is essential for building a successful product in this new paradigm.

Podcast consensus on Jar

Points of consensus

▶Emergent is fundamentally focused on enabling non-technical users to build software, with claims that 80% of its user base has no programming knowledge.Apr 2026

▶The company's core technical strategy revolves around the concept of 'verification,' which they believe is the key to creating autonomous AI agents capable of handling the entire software development lifecycle.Apr 2026

▶A core belief is that the software industry is undergoing a paradigm shift to an 'agent-first' model, and that traditional SaaS companies will not survive without making this pivot.Apr 2026

▶The company extensively 'dogfoods' its own product, using its platform to build internal tools, replace commercial software like Asana, and as the primary coding assistant for its own developers.Apr 2026

Points of debate

▶There is a strategic tension between Emergent's stated focus on serving non-technical users and its public emphasis on highly technical achievements, such as ranking number one on the SWE-bench benchmark.Apr 2026

▶The company predicts foundation models will become commoditized, yet its own strategy relies on nuanced, expert-level application of specific models (Opus, Codex, Gemini) for different tasks and a proprietary 'harness' to boost their performance.Apr 2026

▶A potential operational challenge exists between the company's team structure, which is primarily located in Bangalore, India, and its customer base, which is 70-80% concentrated in the US and Europe.Apr 2026

▶The claim of independently developing concepts like multi-agent systems before their appearance in academic literature positions them as pioneers, but presents a perspective that is inherently difficult to verify and contrasts with the collaborative nature of the AI research community.

Key themes

▶The Agent-First Paradigm ShiftApr 2026

This theme encapsulates the belief that AI agents are not just tools but the future architecture of software itself. Jar predicts that traditional SaaS companies must pivot to an 'agent-first' model to survive, envisioning a future where hundreds of agents collaborate on complex tasks.

This perspective suggests a high-conviction bet on a complete disruption of the existing software market, positioning Emergent not as an incremental improvement but as a foundational platform for the next generation of applications.

▶Democratizing Software Creation for the Non-Technical MajorityApr 2026

A central pillar of Emergent's mission is to empower non-programmers. With 80% of their user base being non-technical, the company's success is tied to abstracting away the complexities of software development, from code generation to production deployment.

Investors should note the massive addressable market this strategy unlocks, but also scrutinize the platform's ability to handle the long-tail of custom requirements that typically demand skilled engineering.

▶Building a Moat Atop Commoditized AIApr 2026

Jar predicts foundation models will become interchangeable commodities. In response, Emergent's strategy is to build durable competitive advantages in the layers above the models, including a proprietary 'agent harness' to boost performance, custom infrastructure, and a focus on the complex 'last mile' of production deployment.

This strategy indicates that the long-term value in the AI stack may not be in the models themselves, but in the specialized orchestration and verification systems that can reliably deploy them for complex, real-world tasks.

▶Pragmatic Engineering and Product Philosophy

Emergent's internal culture emphasizes practical application and user-centricity. This is demonstrated by their intense 'dogfooding' of their own product, a mandatory requirement for all employees to speak with customers weekly, and a focus on solving verification as the key to unlocking agent capabilities.

This hands-on, customer-focused approach, combined with deep technical R&D in areas like verifiers, suggests a culture that is both grounded in solving real-world problems and ambitious in its long-term technical vision.

Source episodes

Timeline

Pre-2023

Prior to co-founding Emergent, Mukund Jar managed a 300-person engineering team at Dunzo in India, while Madhav Jar worked at Zenefits and later started the deep learning team at Amazon.

Late 2023

The team applied to Y Combinator with the initial idea of automating software testing.

Early 2024 (Implied)

Pivoted from software testing to automating the entire software engineering process, based on the key insight that solving for 'verification' was the critical enabler for autonomous agents.

2024

Within two months of development, Emergent's coding agent achieved the number one ranking on the SWE-bench benchmark, a significant technical milestone.

Present

Emergent now serves a global user base, 80% of whom are non-technical, and has seen a customer company raise $4 million in funding based on an application built on its platform.

Suggested prompts

How does Emergent's focus on non-technical users align with its promotion of highly technical achievements like the SWE-bench ranking, and who is the primary audience for each message? &nearr;What are the specific mechanisms within Emergent's 'proprietary agent harness' that allegedly extract 20-30% more performance from foundation models? &nearr;Given the prediction that SaaS companies must become 'agent-first,' what are the primary barriers to this transition for incumbents, and how does Emergent's architecture address them? &nearr;How does Emergent's strategic decision to build its own infrastructure, including a custom Kubernetes stack, impact its unit economics and scalability compared to competitors using third-party sandbox providers? &nearr;

Key concepts

Agent-First Companies 1 ep Multi-Agent Architecture 1 ep Software Verification 1 ep Non-Technical Users 1 ep Foundation Model Commoditization 1 ep Continual Learning & Agent Memory 1 ep SWE-bench Benchmark 1 ep Dogfooding 1 ep Proprietary Agent Harness 1 ep Go-to-Market Strategy 1 ep

Notable quotes

“unless your SaaS company pivots into like an agent first company uh you know you know I think that going to be hard to sort of survive”

Mukund or Madhav Jar · AI Is Unlocking Millions Of New Builders

“today 80 of users who are on the platform are non users with zero programming knowledge”

Mukund or Madhav Jar · AI Is Unlocking Millions Of New Builders

“i think by end of the year you'll have you know agents which are running 24 hours and like maybe hundreds of agents collaborating on just single task.”

Mukund or Madhav Jar · AI Is Unlocking Millions Of New Builders

“we built Soda Coding Agents, which became world number one on Sweebench in two months of time.”

Mukund or Madhav Jar · AI Is Unlocking Millions Of New Builders

Report last updated: Apr 28, 2026

Get started free

Back to Entities Intelligence Report