Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Join the event that trusts business leaders for almost two decades. VB Transform brings together people who build a real business AI strategy. Learn more
ZencoderThe artificial intelligence coding startup founded by serial entrepreneur Andrew Filev, announced today the public beta launch of AmountAn agent powered by AI designed to automate end -to -end software tests. This critical but often slow step can delay product versions in days or weeks.
The new tool represents the latest attempt to zencoder to distinguish itself on the market from the increasingly congested AI coding assistants, where companies rush to automate not only code generation but workflows for development of whole software. Unlike existing AI coding tools which focus mainly on code writing, Zenteter targets the verification phase – ensuring that the software works as expected before reaching customers.
“Verification is the missing link in the scaling of development based on the AI of experimentation in production,” said Filev in an exclusive interview with Venturebeat. The CEO, which previously founded a project management company Section and sold it to Citrix for $ 2.25 billion In 2021, added: “Zenteter does not only generate tests – he gives developers the confidence necessary to send by validating that their code generated by AI or man does what he is supposed to do.”
The announcement comes as the AI coding market undergoes rapid consolidation. Last month, Zencoder acquired machineAnother AI coding assistant with more than 100,000 downloads. At the same time, Openai concluded an agreement to acquire the Windsurf Coding Tool for About $ 3 billion (The agreement was concluded in May). The movements emphasize how companies rush to build full AI development platforms rather than occasional solutions.
Why the software tests have become the largest road road dam powered by AI
Amount Takes a persistent challenge in the development of software: long feedback loops between developers and quality insurance teams. In typical corporate environments, developers write code and send it to QA teams for tests, often waiting for several days for comments. Until then, developers have passed to other projects, creating an expensive change of context when the problems are discovered.
“In a typical engineering process, after a developer has created a functionality and sends it to QA, they receive comments several days later,” Filev told Venturebeat. “Until then, they have already moved on to something else. This context of change of context and back and forth – in particular painful during the cracking in liberation – can stretch simple fixes in the tests of a week.”
Early customer Club solutions group Spectacular improvements have pointed out, CEO Mike Tervino declaring: “What took our QA team for a few days now take the developers for 2 hours.”
The moment is particularly relevant because the AI coding tools generate increasingly important volumes of code. While tools like Github co -pilot And Cursor have accelerated the generation of code, they have also created new quality insurance challenges. Filev believes that if AI tools increase the generation of code of 10x, test requirements will also increase by 10x – of crushing traditional QA processes.
How the AI agents of Zenteter click on the buttons and fill out forms like human testers
Unlike traditional test frames who force developers to write complex scripts, Amount Works on simple English instructions. The AI agent can interact with applications such as a human user – click buttons, fill out forms and navigate in software workflows – while validating both frontal user interfaces and backend features.
The system fits into existing test frames, including playwright and selenium, rather than replacing them entirely. “We absolutely do not like people who abandon things that are part of our DNA,” said Filev. “We believe that AI should take advantage of the processes and tools that already exist in industry.”
Zenteter offers five basic capacities: quality tests led by developers during the development of functionality, acceleration of the AQ for a complete creation of full tests, improvement of quality for the code generated by AI, automated test maintenance and autonomous verification in continuous integration pipelines.
The tool represents the latest addition to the wider multi-agent platform of Zencoder, which includes coding agents to generate software and unit test agents for basic verification. Society “Repo gropmingThe technology Analysis of whole code standards to provide a context, while a error correction pipeline aims to reduce the bugs generated by AI.
The launch intensifies competition on the AI development tools market, where players established like Microsoft Github co -pilot and new entrants like Cursor are in the running for the Mindshare developer. Zencoder’s approach to the construction of specialized agents for different development phases contrasts with competitors focused mainly on code generation.
“At this stage, there are three solid coordination products on the market that are production note: it is us, cursor and Windisurf,” said Filev in a recent interview. “For small businesses, it becomes more and more difficult to compete.”
The company requests higher performance on industry references, declaring success rates of 63% on Swe-Bench checked tests and around 30% on the new Swe-Bench Multimodal Benchmark – Filev results indicates the better previous performance.
Industry analysts note that automation of end -to -end tests represents a next logical step for AI coding tools, but successful implementation requires a sophisticated understanding of the application logic and user workflows.
What business buyers must know before adopting AI test platforms
Zencoder’s approach offers both opportunities and challenges for corporate customers assessing AI test tools. Society SOC 2 Type II,, ISO 27001 And ISO 42001 Certifications address essential safety and compliance problems for large organizations.
However, Filev recognizes that the prudence of the company is justified. “For companies, we do not completely recommend changing the life cycles of software development,” he said. “What we recommend is a-increase, where they can now have reviews of revision and acceptance of fast Ai code which reduce the amount of work that must be carried out by the next part of the pipeline.”
The company’s integration strategy – working in existing development environments such as Visual Studio code And Jet-Brains Ides Rather than obliging platform switches – can call on companies with established tool channels.
The race to automate it the development of software from the idea to the deployment
Zenteter Zencoder’s launch positions to compete for a larger share of software development work as the AI tools develop beyond the generation of simple code. The company’s vision extends to the complete automation of requirements to the deployment of production, although Filev recognizes current limitations.
“The next jump will be production requirements – all,” said Filev. “Can you now kill him so that you can have natural language requirements, then AI could help you break down, build an architecture, build code, create a revision, check it and ship it in production?”
Zencoder offers ZenSestester via three price levels: a free basic version, a $ 19 business plan per user and per month and a business option of $ 39 per month per month with premium support and compliance features.
For an industry that always wonders if artificial intelligence will replace programmers or make them simply more productive, Zenteter suggests a third possibility: AI that manages tedious verification work while developers focus on innovation. The question is no longer whether the machines can write code – it is whether you can trust it to test it.
Source link
