Applications & platforms
Web
iOS
Android
Windows
Mac
Linux
Cloud
On-premise
IoT / edge
Technologies
Software, AI & infrastructure technologies
Our stack covers classic business applications, web/mobile/desktop platforms, AI systems, real-time voice, document search, IoT and production operations. The point is not to pile up buzzwords: it is to pick the right building blocks to ship quickly, maintain for years and keep technical control.
Back-end
Ruby on Rails, Python, FastAPI and Node.js depending on context. REST/GraphQL APIs, WebSockets, workers, queues and enterprise integrations.
Front-end
TypeScript, HTML/CSS, Tailwind, React/Next.js, Hotwire, Turbo and Stimulus. Accessibility, performance and design-system discipline.
Data & AI
LLMs, embeddings, RAG, vector search, neural networks, deep learning, computer vision, audio, evaluation and MLOps.
Audio & realtime
Audio engineering, realtime stream processing, streaming, WebSockets, SSE, WebRTC, STT/TTS and low-latency pipelines.
Databases
MySQL, PostgreSQL, Redis, pgvector, indexing, hybrid search, migrations, retention and performance according to context.
Cloud & DevOps
Docker, CI/CD, Heroku, AWS, GCP, Cloudflare, GitHub Actions, logs/metrics/traces and SRE practices tuned to the product stage.
Security
Hardening, secrets, IAM/RBAC, SSO, dependency audits, SAST/DAST, GDPR and operational evidence.
Stack used day to day
The building blocks we know how to assemble.
These technologies cover Osmove projects including PhoneVoice, Twoody, Backlog and Lint, as well as client work. The final selection always depends on the business need, team, data, infrastructure and risk level.
Languages & runtimes
Ruby
Python
Node.js
TypeScript
JavaScript
SQL
HTML
CSS
Application frameworks
Ruby on Rails
React
React Native
Next.js
Tailwind CSS
Hotwire
Turbo
Stimulus
APIs, servers & realtime
FastAPI
Rails API
Node.js APIs
REST
GraphQL
WebSockets
ActionCable
Server-Sent Events
Twilio Media Streams
AI, ML & deep learning
LLMs
RAG
Embeddings
Vector search
Neural networks
Deep learning
TensorFlow
Keras
PyTorch
scikit-learn
Audio, voice & signal
Audio engineering
Realtime audio
Streaming
WebSockets
SSE
WebRTC
Librosa
SciPy
NumPy
STT
TTS
OpenAI Realtime API
VAD
MFCC
Data & search
MySQL
PostgreSQL
Redis
SQLite
pgvector
OpenSearch / Elasticsearch
Vector indexes
Hybrid search
Reranking
Data pipelines
Infrastructure & operations
Docker
Heroku
AWS
GCP
Cloudflare
GitHub Actions
CI/CD
MLOps
Observability
Security & quality
OAuth
JWT
SSO
RBAC
Secret management
SAST/DAST
RuboCop
ESLint
Ruff
Biome
Stylelint
Technology choices
No imposed stack: choices guided by the project.
We prefer an architecture your teams can understand, operate and evolve. The right technology depends as much on your existing systems as on the product being built.
Existing systems
Compatibility with your databases, APIs, ERP, CRM, hosting constraints and internal standards.
Data & search
Relational, document, cache, full-text search, vectorization or hybrid retrieval depending on volumes and usage.
Team & operations
Available skills, documentation, monitoring, run cost, backups, migration and maintenance.
AI when it serves the workflow
LLMs, ML, vision, audio or agents only when user value and guardrails are clear.
Stacks by product
Technologies behind our own projects.
We talk about technologies we use in real products. That experience then informs our choices on client projects.
Voice AI
PhoneVoice
AI phone agents, real-time audio, inbound/outbound calls and integration into business workflows.
Ruby on Rails
Twilio Voice
Media Streams
OpenAI Realtime API
WebSockets
ActionCable
PostgreSQL
STT/TTS
Prompt engineering
AI assistant
Twoody
Voice-first AI assistant spanning web, calls, email, agents, commercial cloud and local self-hosted edition.
Rails
TypeScript
Realtime voice
WebRTC
OpenAI Realtime API
Multi-LLM
RAG
MCP
Python
Raspberry Pi / Jetson
Cloud / self-hosted
Development agents
Backlog
Plan, run, claim, workspace, CLI/SDK and workflow orchestration for AI coding agents.
Ruby on Rails
TypeScript
PostgreSQL
Queues
CLI/SDK
OAuth
Osmove Identity
Token exchange
GitHub Actions
Observability
Code quality
Lint
Universal linter and AI review CLI to normalize code quality across ecosystems.
Node.js
TypeScript
npm
CLI
ESLint
RuboCop
Ruff
Biome
Stylelint
AI review
CI
Frontend & backend foundations
We favor technologies that maximize delivered value while reducing unnecessary complexity. Ruby on Rails, Python, FastAPI and Node.js are our backend foundations: high productivity, mature ecosystems and strong tooling for testability and security. On the frontend, we use TypeScript, Tailwind, React/Next.js or Hotwire depending on the expected level of interactivity, SEO, realtime behavior and long-term maintenance.
Data & AI in production
For data and AI, we activate the right building blocks for each use case: LLMs, embeddings, RAG, vectorization and hybrid search for document corpora, TensorFlow, Keras, PyTorch and scikit-learn for machine learning and deep learning, Librosa, SciPy and NumPy for audio and signal processing. Productionization follows an MLOps framework: evaluation, monitoring, cost control, confidentiality, traceability and fallbacks.
Audio, streaming & realtime
We can handle the audio chain end to end: capture, cleanup, segmentation, VAD, feature extraction, stream processing, STT/TTS and low-latency orchestration. Realtime interfaces rely on WebSockets, Server-Sent Events, WebRTC or media streams depending on the need: voice agents, AI phone calls, live transcription, speech synthesis, audio monitoring or UI synchronization.
Databases & Cloud/DevOps
Databases are selected and engineered for robustness, scalability and compatibility with your information system: MySQL, PostgreSQL, Redis, SQLite, full-text search, vector indexes or dedicated engines depending on the need. On Cloud/DevOps, we rely on Heroku, AWS, GCP, Cloudflare, Docker and GitHub Actions, with reproducible CI/CD pipelines and usable observability: logs, metrics, traces, alerting and rollback procedures.
Security & user experience
Security is pervasive: secret management, hardening, RBAC/SSO, dependency audits, SAST/DAST scans, backup policies and restore testing. We align choices with your requirements (including GDPR) and schedule regular reviews. Finally, we incorporate user‑experience constraints (accessibility, web performance, design systems) to ensure a high perceived quality on the web and on iOS/Android.