Join our daily and weekly newsletters for the latest updates and the exclusive content on AI coverage. Learn more
Zencoder announced today the launch of Zen agentsA platform that allows the creation and sharing of specialized AI tools for the development of software. The version includes a Market Open Source Where developers can contribute and discover personalized agents, marking a significant change in the way development teams take advantage of artificial intelligence.
While existing AI coding assistants have mainly focused on stimulating the developers’ individual productivity, Zencoder’s approach addresses the collaborative reality of modern software engineering, where delays often occur between coding and feedback loops.
“If you look at the tools used today for a real engineering AI, these are mainly coding agents with an IDE,” said Andrew Filev, CEO and founder of Zencoder, in an exclusive interview with Venturebeat. “And if you dig a single deeper layer, you will see that they are generally focused on the individual developer. Everything is perfectly logical, because it all starts with the developer, right?”
But Filev points to a critical gap in current solutions: “There is all this layer of things that you can do beyond individual engineers, because engineers do not work alone. In a successful software company, development occurs in teams. ”
How Zen agents reduce development cycles by automating intermediate steps
The new platform addresses this difference by allowing teams to create and deploy personalized agents adapted to frames, workflows or specific code bases. These agents can be shared between organizations, ensuring coherent practices while eliminating repetitive tasks.
What technically distinguishes Zen agents is its implementation of Model context protocol (MCP)a standard from anthropic and supported by OPENAI This allows large language models to interact with external tools.
“As part of this launch, we present our own register which has more than 100 MCP servers,” said Filev. “We have created this because there is not yet a standard register available. If a standard register existed, we simply connect to it, because our real value comes from our agents and our specialized tools. ”
Industry analysts see this as a natural progression of development tools. The initial wave of AI coding assistants has provided immediate productivity increases for individual tasks, but has failed to approach the collaborative nature of business software development, where time is often lost in transfers between team members.
Zen agents aim to approach these transfers by allowing specialized agents to automate parts of the development life cycle, from the examination of the test to tests. “For example, let’s say you have an agent who examines the code,” said Filev. “Imagine that there is an agent in whom you trust. The agent does not even necessarily need to be as good as a human, because if he finds problems and provides comments immediately, you can solve these problems immediately.”
The platform is designed to be ready for the company, with zencoder touting ISO 27001,, SOC 2 Type II certificationAnd ISO 42001 For systems responsible for the management of AI – the references necessary for adoption in organizations concerned with security.
The most distinctive aspect of the launch may be the Market Open SourceThis allows the larger developer community to contribute specialized agents. This approach reflects successful open source ecosystems as Visual Studio code extensions Or NPM packageswhere community contributions considerably expand capacities beyond that a single supplier could develop.
“I am a great supporter of collective intelligence,” noted Filev. “There are so many use cases that we have not even thought about yet, and even if we imagine them all, we would never have the resources to cover them ourselves.”
The first adopters have already found the value of the creation of specialized agents. “I was impressed by the examples that integrate several stages in their workflow,” said Filev. “For example, you can withdraw a Wireframe from Figma, automatically generate code according to it, then submit a traction request – as well as a transparent process.”
Another notable example meets accessibility requirements – an area often recognized as important but frequently decompressed in tight deadlines. “Our developer defender has created an agent who improves the accessibility of the code,” said Filev. “Everyone in the software agrees that accessibility is extremely important, but in reality, the teams do not always have time to meet these needs correctly.”
According to Matt Walker, co-founder and CTO of Simon datawhich was cited in the press release, the impact has been measurable: “Zen agents mark an important development in AI development.
Beyond the coding: the breed to the state of flow of the improved developer AI-Ai
Pricing Zen agents Currently follows a structure on several levels. “Our price plans are simple: we offer a free level, as well as monthly options of $ 20 and $ 40,” said Filev, although it has noted that the use grew, the company is considering extended options. “The way I think about it is simple – the more you use it, the more money you save.”
For the future, Filev sees Zen agents evolve towards greater autonomy, not to replace engineers but make them considerably more productive. “We run towards autonomy-not with the aim of replacing engineers, but by the vision of making engineers 10 times more productive,” he said.
This vision extends beyond writing code to maintain what developers call “the state of flow” – periods of uninterrupted and very productive work. “Our company has Zen on his behalf, and it is not productive to start working on something, then jumping for something else, to come back to the original task later,” said Filev. “If we can keep you in this state of flow, then mission accomplished, right?”
Although Zencoder is initially focused on software engineering applications, Filev alluded to a wider potential. “Many of my technology friends already use this technology for engineering purposes,” he said, mentioning personal assistants and marketing automation as examples. “I am curious to see what the community creates with it – there is a real possibility that it can gain ground in a much wider context.”
As AI tools mature in the software development space, Zen agents point to a future where technology becomes less on the replacement of individual tasks and more on the orchestration of the entire life cycle of development. By focusing on the spaces between the developers-rather than the developers themselves-Zencoder may have found the way to this elusive “Zen” state that each coder is looking for: the construction of a software that has the impression that he is writing himself.
