Teen founders skip graduation to join SF accelerator, building AI efficiency startup

·1 views

James Yang and Anish Paleja, both 18-year-old Canadians from Ontario, skipped their high school graduation to attend Founders Inc.'s Off Season II accelerator in San Francisco. The two met at a Major League Hacking hackathon at age 15, when neither could code nor had any product development experience. They have since founded Brevitas, a startup focused on reducing token waste in multi-agent AI systems to help companies cut costs on AI workflows. Despite the program being only six days old, the pair say they have already achieved roughly six months' worth of progress. Both are prolific hackathon participants, with Paleja having attended 36 and Yang 14 in the past year alone.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Docusign Deploys NVIDIA Nemotron Parse to Extract Complex Contract Tables at Scale

Docusign has integrated NVIDIA's Nemotron Parse, a vision-language model built for document understanding, into its production document processing pipeline to improve contract table extraction. The company, which handles millions of daily transactions across nearly 1.9 million customers in over 180 countries, has long struggled with accurately pulling structured data from complex contract tables containing merged cells, multi-page layouts, and nested formatting. General-purpose AI models failed to reliably handle these structures, often requiring costly manual corrections that slowed enterprise workflows. The solution was unveiled at the AI Engineer World's Fair in San Francisco, where Docusign and NVIDIA jointly demonstrated the system's architecture. The pipeline uses vLLM to serve the model while keeping all sensitive agreement data within Docusign's secure, on-premises environment to meet strict data governance requirements.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer builds typed, API-first CMS for business data using PHP, OpenAPI, and MCP

A developer has built NeNe Records, an open-source, API-first content management system designed to handle structured business data through typed schemas and documented APIs. Unlike WordPress, which uses flexible but loosely typed metadata storage, NeNe Records enforces field types — such as text, enum, image, and relation — at the API level, not just in the frontend UI. The system is built on PHP 8.4 using a custom framework called NENE2, with a React and TypeScript admin interface and an OpenAPI 3.1 contract. AI integrations are supported via an MCP tool catalog derived directly from the documented API, with a strict rule that AI tools must interact through the application API rather than the database. The project, part of a broader suite of self-hosted tools for small teams in Japan, is still in active development but already supports multi-tenant organization scoping, media handling, and public pages.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Puppet Enterprise 2025.11 Adds PostgreSQL CA Storage and Patches 60 CVEs

Puppet has released Puppet Enterprise 2025.11, a major update featuring an optional PostgreSQL-backed CA storage system that replaces file-based storage for improved performance and reliability. The release upgrades the bundled PostgreSQL from version 14 to 17 and integrates GPT-5 series models into the Infra Assistant for better query responses. Advanced patching capabilities have been expanded, including a new puppet_run_concurrency setting and improved cron scheduling validation. Platform support has been updated to include macOS 26 on both ARM and x86_64 architectures, while Ubuntu 18.04 and 20.04 have been dropped. Nearly 60 CVEs were resolved in this release, and a companion patch for the LTS version PE 2023.8 has also been published simultaneously.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Why AI Coding Agents Need to Be Collaborative, Not Solo Tools

Sergey Karayev, co-founder of Superconductor, argues that current AI coding agents like Claude Code, Cursor, and Codex are built for individual users, creating coordination problems when used by teams. At Superconductor, the team shifted to cloud-based agent sessions that any teammate can join, giving everyone access to the full history of decisions, dead ends, and ongoing work. This approach improved code reviews, simplified handoffs, and allowed teams across time zones to maintain continuous progress without losing context. Beyond coding, the company also deploys agents in meetings and for tracking team goals. Karayev believes the future of collaborative work lies in shared AI sessions rather than isolated, per-person chat windows.

0 comments Read more at DEV Community