FlossWare

OPEN SOURCE · AI · ENGINEERING

Software systems built to stay understandable.

FlossWare is a free-first engineering ecosystem for modular infrastructure, AI-assisted systems, distributed services, and reusable open-source foundations.

Explore GitHub Explore the architecture

Openstandards first

Modularloosely coupled systems

AI-readyREST + MCP + events

Documenteddecisions preserved

What is FlossWare?

FlossWare builds reusable engineering foundations and reference implementations using open standards, explicit configuration, and loosely coupled architectures. The goal is deliberately unfashionable: make powerful systems understandable, replaceable, and maintainable.

Architecture

Service boundaries, orchestration, routing, distributed fleets, consensus, and stable integration contracts.

Knowledge systems

Scraping, chunking, embeddings, vector search, graphs, and the pipelines connecting them.

Open-source projects

Reusable libraries, infrastructure tooling, engineering standards, and reference implementations.

Architecture

Configuration is the source of truth.Behavior should be explicit rather than hidden in code.

Minimal by default.Capabilities are enabled deliberately.

Contracts over coupling.Stable interfaces let components evolve independently.

Open standards.Prefer interoperable designs over vendor lock-in.

                         Clients / Agents
                                |
                   REST APIs / MCP Tool Interfaces
                                |
                         Service Boundaries
                         /               \\
                        /                 \\
              Stored Procedures       Message Bus
                    |                       |
                    v                       v
                Databases             Event Consumers

REST provides synchronous service contracts.
MCP provides AI agent and tool integration contracts.
Message buses provide asynchronous workflows and event-driven integration.
Stored procedures provide database abstraction where they add value.
ADRs preserve architectural decisions.

Architecture documentation

Orchestration Layer — REST API, task classification, request lifecycle
Distributed Fleet — heterogeneous cluster topology and SSH-based distribution
Consensus Engine — multi-model synthesis and arbiter patterns
Model Routing — Thompson Sampling, capability matrix, fallback chains

Knowledge systems

Ingest

Web scraping and document acquisition from diverse knowledge domains.

Transform

Semantic chunking and preparation for downstream retrieval.

Retrieve

Vector embeddings, HNSW indexing, and similarity search.

Connect

Knowledge graphs and relationship traversal across stored knowledge.

Knowledge documentation

Data and infrastructure

System	Role
PostgreSQL + pgvector	Relational storage and vector similarity
Redis	Pipeline queues, caching, and rate limiting
OrientDB	Graph storage and relationship traversal

Learning and optimization

The engineering work also includes experimental systems for adaptive decision-making and configuration evolution.

Thompson Sampling — Bayesian multi-armed bandit for strategy selection
Genetic Algorithms — configuration evolution across multiple domains

Operations

Deployment — Ansible-automated fleet deployment
Monitoring — fleet health and provider tracking
Scaling — horizontal and vertical scaling and bottleneck identification

Projects

engineering-standardsArchitecture decisions, engineering principles, and development standards. commons-javaShared Java foundation libraries. tftp-osReference implementation for infrastructure provisioning and automation.

Development

About FlossWare

FlossWare is an open-source engineering effort led by Scot P. Floess (Flossy). The work spans distributed systems, enterprise Java, search, AI/ML, infrastructure automation, and open-source tooling.

For the professional résumé and background, visit sfloess.github.io. For the Salesforce deployment project, visit Solenopsis.