B3dmar builds AI-native developer tools and SaaS products that operate in production, not just demos. Its product line includes 3ngram, which gives AI assistants persistent memory via MCP; GTM:OS, a markdown-based go-to-market system; and b3dmar-auth-core, a production-tested authentication library. The company emphasizes open source, shared libraries, and practical workflows used daily by real users.
Funding
Funding not disclosed
Founders
Product
Problem
AI assistants lack persistent memory, forcing users to repeatedly provide context across sessionshare decisions, deadlines, and project history are lost, compounding friction with every use. Additionally, solo founders and consultants lack affordable go-to-market tools that replace expensive CRM suites with simple, intelligent workflows.
Solution
B3dmar ships a suite of AI-first products designed to solve these problems. 3ngram connects to AI assistants via the Model Context Protocol (MCP), giving them a persistent memory of decisions, commitments, and context, with features like semantic recall, morning briefings, and structured accountability. GTM:OS turns a markdown file into a go-to-market operating system, providing slash commands and agents for pipeline management, lead qualification, and personalized outreach. b3dmar-auth-core is a production-tested authentication library extracted from B3dmar's own services, offering JWT handling, password hashing, RBAC, and token revocation. All products are built with AI at the architecture levelahanostic of AI as an afterthought, and are designed to ship and run in production, not merely demo.
Target Audience
Primary users are solo founders, consultants, and small engineering teams building AI-powered workflows, as well as developers of FastAPI services needing production-tested authentication and persistent AI context.
Features
- 3ngram provides 30+ MCP tools for AI assistants, including remember/recall with semantic search, commitment tracking, automatic morning briefings, and provider handoff in JSON
- 3ngram's backend uses FastAPI, PostgreSQL with pgvector for embeddings, and OAuth 2.0 with JWT auth, and is compatible with Claude Code, Cursor, and Windsurf
- GTM:OS installs as a CLI with markdown-based skills, reading a pipeline.md file, and runs entirely in the user's AI context windowhiwith no external APIs or databases
- GTM:OS ships with 8 slash commands and 3 agents (Researcher, Copywriter, Analyst) for pipeline analytics, prospect research, and meeting prep
- b3dmar-auth-core includes Argon2id password hashing, Redis-backed token revocation, rate limiting, and Pydantic v2 schemas, with a stateless, domain-agnostic design