Urgent.News

What's breaking now, across thousands of outlets.

Tech

ForgeCode’s Tushar Mathur wants software development to run like a factory, not a project

At TechSparks 2026, ForgeCode Founder and CEO Tushar Mathur demonstrated how AI agents can manage software development from specification to deployment, while engineers focus on supervision and decision-making.

ForgeCode’s Tushar Mathur wants software development to run like a factory, not a project

Tushar Mathur, the founder and CEO of ForgeCode, is challenging the traditional way of scaling software development. Instead of simply adding more engineers, time, and resources, Mathur believes that AI is transforming the software development process. While AI has made it cheaper and quicker to write code, the real bottleneck now lies in reviewing, testing, and deploying the generated code.

ForgeCode aims to address this issue by creating an AI-native software factory. This platform utilizes AI agents to handle the entire software development lifecycle, from specifications to deployment, with human oversight for decision-making and feedback. At the recent TechSparks 2026, Mathur showcased how the platform operates and emphasized the importance of reviewing, testing, and deploying code in the current software delivery landscape.

As AI increases individual productivity, the bottleneck has shifted from writing code to managing the processes that follow. The AI-native software factory aims to keep software development moving forward without requiring engineers to manage every individual task. Mathur outlined six principles that he believes are crucial for building an effective AI-native software factory, including autonomous operation, high scalability, cost-effectiveness, measurability, self-improvement, and business-driven development.

This approach seeks to make software development more autonomous while maintaining alignment with business priorities. When it comes to trade-offs and non-functional requirements, Mathur stated that decisions that can be objectively verified can be handled by the software factory, which continuously tests outcomes and optimizes whenever a check fails.

However, complex decisions requiring human judgment, such as scalability and database architecture, still fall under the purview of engineers and business leaders. By learning from every task, decision, and correction, the AI-native software factory improves its processes over time and operates with greater autonomy.

Written by urgent.news from YourStory's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at yourstory.com →

More in Tech

More from Wednesday 4 November →