What happened
Tessl, an AI-native software development firm, launched an open standards platform for building 'software factories' in New York. The platform is designed to shift software development from individual developers using coding assistants to systems where AI agents handle a greater share of the development lifecycle. It offers an agent-agnostic and model-agnostic approach, connecting ticket implementation, code review, and vulnerability remediation. The system allows engineering teams to define reusable 'skills' and combine them into 'loops' that include verification and feedback mechanisms. Tessl reported internal metrics showing that 80% of pull requests required no human hand-editing and that the cost per pull request dropped from $75 to $13. The launch includes a promotional offer of $200 in usage credits for new teams.
Tessl, founded in 2024 by Snyk founder Guy Podjarny, has launched an open standards platform for building software factories. The company has raised over $120 million to date. The platform is positioned as an agent-agnostic and model-agnostic solution, meaning it is not tied to specific AI models or agent frameworks.
The core functionality of the platform involves connecting ticket implementation, code review, and vulnerability remediation. Engineering teams can define reusable 'skills' that dictate how agents perform specific tasks. These skills are combined into repeatable workflows called 'loops,' which are supplemented with verification and feedback mechanisms to check the resulting work. The design intends for these loops to become self-improving over time based on the feedback inputs.
Tessl cited concerns over poor-quality 'AI slop' outputs, security risks from increased agent access, and unpredictable costs as factors slowing the adoption of agent delegation. The platform aims to address these by providing a structured environment for delegation. Ready-to-configure workflows include 'Ticket to PR' for implementing scoped issues, 'Code Review' for checking changes against standards, and 'Vulnerability Remediation' which uses scanners like Snyk to identify and propose fixes.
The company reported internal usage statistics indicating that 80% of a week’s pull requests required no human hand-editing when handled by the platform. Additionally, Tessl stated that the cost per pull request was reduced from $75 to $13 across implementation, review, security fixes, and other workflow steps. These figures are self-reported by Tessl and have not been independently verified by third parties.
The launch includes a promotional offer of $200 in Tessl usage credits, valid for 30 days from activation, available until November 30th for teams trying the platform for the first time. Guy Podjarny emphasized that the challenge is not assigning work to agents, but getting them to understand boundaries and when to stop, positioning verification as a fundamental component of increased agent autonomy.
Why it matters
This launch addresses specific industry pain points regarding the quality, security, and cost unpredictability of AI-generated code, often referred to as 'AI slop.' By providing a structured, verifiable framework for agent delegation, Tessl aims to make autonomous software development more reliable for enterprise teams. The reported cost reduction and reduction in manual editing suggest a practical shift in how engineering teams might integrate AI agents into their daily workflows, moving beyond simple code completion to full lifecycle management. This is significant for the AI industry as it represents a move toward standardized, composable infrastructure for AI agents in software engineering, potentially setting new benchmarks for efficiency and quality control in AI-assisted development.
The launch is significant because it targets the transition from AI-assisted coding to AI-automated software development. By providing an open standards platform, Tessl aims to create a common infrastructure for 'software factories,' which could standardize how AI agents are deployed in engineering teams.
The reported cost reduction from $75 to $13 per pull request, if accurate, represents a substantial economic incentive for enterprises to adopt agent-based workflows. This could accelerate the shift away from traditional human-centric development models toward hybrid or fully automated pipelines.
The emphasis on verification and feedback mechanisms addresses a key criticism of current AI coding tools: the lack of reliability and the potential for introducing security vulnerabilities. By integrating tools like Snyk for vulnerability remediation, Tessl attempts to mitigate the security risks associated with autonomous code generation.
As an agent-agnostic and model-agnostic platform, Tessl's approach may allow enterprises to switch between different AI models or agents without re-architecting their workflows. This flexibility could be a competitive advantage in a rapidly evolving AI landscape where model capabilities change frequently.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
crm_get_transaction(id='4092').An agent must create a draft calendar event for Tuesday at 2 p.m. Which evidence would establish the requested result?
What to watch next
Monitor independent verification of Tessl's internal performance metrics, particularly the 80% reduction in manual editing and the cost savings claims. Watch for enterprise adoption rates and whether the 'open standards' aspect leads to broader industry integration or remains proprietary in practice. Additionally, observe how the platform handles security vulnerabilities in agent-generated code, as this is a critical concern for enterprise deployment.
Independent verification of the 80% reduction in manual editing and the cost savings claims is crucial. These metrics are self-reported by Tessl and may not reflect typical enterprise environments or diverse codebases.
The adoption of the 'open standards' aspect of the platform will determine its long-term impact. If other vendors or open-source projects adopt these standards, it could become a de facto standard for AI software factories. If not, it may remain a proprietary solution despite the 'open' branding.
Security implications of agent-generated code will be a key area of scrutiny. While Tessl integrates vulnerability scanners, the effectiveness of these tools in catching subtle or novel vulnerabilities introduced by AI agents will need to be demonstrated in real-world scenarios.
The competitive landscape for AI coding agents is intensifying, with major tech companies and startups launching similar tools. Tessl's ability to differentiate itself through its focus on verification and cost efficiency will be critical to its success.