Full-Stack Engineer (.NET/React)

About Synthreo

Synthreo is the AI-native platform for Managed Service Providers. We give MSPs everything they need to launch an AI-as-a-Service business without hiring AI experts — from our no-code agent builder (Builder) to a white-label ChatGPT-style portal (Threo) that their clients actually use.

We’re a small, senior team shipping real products that MSPs deploy to thousands of end users. Our platform handles secure RAG pipelines, multi-tenant isolation, and deep ConnectWise PSA integrations — the kind of infrastructure that lets our partners monetize AI without becoming AI companies themselves.

The Role

You’ll work across three core products: Threo (our client-facing AI portal with real-time SignalR streaming), Builder (our no-code agent platform with drag-and-drop workflow UI), and Tenant Management (our admin console for authentication, customer management, billing, secrets management, and access control).

This role is backend-heavy (~70–80%) with meaningful frontend work. You’ll design APIs and data models, then build the React UIs that consume them. No pixel-perfect design work — we need someone who can ship functional admin interfaces, data grids, forms, and real-time streaming UIs. You’ll work directly with founders and a senior team; expect high ownership and real impact from day one.

What You’ll Do

Backend (~70–80%)

• Design and ship services in .NET 6–9/C# (ASP.NET Core, Minimal APIs, background workers/hosted services)
• Own data modeling in PostgreSQL — schema design, Dapper queries, DbUp migrations, indexing, query optimization
• Work with Apache Cassandra 5.0 for agent and conversation data (wide-column design, partition/clustering keys, native vector search with SAI)
• Build and extend Tenant Management: authentication flows (JWT, OAuth 2.0, Azure AD), RBAC, secrets management, billing integration, and cross-tenant access rules
• Build REST APIs with idempotent workflows, back-pressure, and retry strategies; integrate with RabbitMQ for async pipelines
• Implement SignalR hubs for real-time chat streaming and live updates
• Integrate with LLM providers via Azure AI Foundry SDK and AWS Bedrock SDK — implement function calling, manage token streaming, handle retries and fallbacks
• Collaborate with our Python agent execution engine — understand how agents are orchestrated, how tool calls flow between .NET and Python, and debug cross-service issues

Frontend (~20–30%)

• Build and iterate on React 19 interfaces — admin consoles, data grids, forms, workflow builders
• Implement real-time features using SignalR client for chat streaming and live data updates
• Work with DevExtreme components for data grids and UI elements; React Flow for diagram/workflow visualization
• Write clean component code with React Hooks (useState, useEffect, useCallback) and Context API for state management

Infrastructure

• Ship via Docker and GitHub Actions to Azure Container Apps; contribute to Terraform infrastructure as needed

What We’re Looking For

We hire based on skills, curiosity, and growth potential — not years of experience. If you’re excited about AI and can ship code, we want to talk.

You Have

• Solid experience building production systems in .NET/C# — you’ve shipped APIs and worked with data layers
• PostgreSQL experience: schema design, writing queries, understanding indexes. Experience with Dapper or similar micro-ORMs is a plus
• Solid React experience: hooks, component architecture, state management. You can build functional UIs without hand-holding
• Strong grasp of Design Patterns and SOLID principles — you write clean, maintainable code on both frontend and backend
• Distributed systems understanding: concurrency, async/await patterns, transactions, and consistency trade-offs
• JWT authentication implementation on both server and client side
• CI/CD and Docker fluency
• LLM integration experience: working with AI model APIs (Azure OpenAI, AWS Bedrock, or similar), function/tool calling patterns, and token streaming
• Clear async communication skills — critical for remote collaboration

Bonus Points

• Azure AI Foundry SDK and/or AWS Bedrock SDK experience
• Python proficiency — ability to read, debug, and collaborate on our Python agent execution engine
• Apache Cassandra experience (wide-column modeling, SAI indexes, native vector search)
• SignalR — both hub implementation and client consumption
• DevExtreme React components or similar enterprise UI libraries
• React Flow, JointJS, or similar diagram/workflow libraries
• RabbitMQ or similar message queue experience
• Azure experience (Container Apps, Azure AD, Azure OpenAI)
• Microsoft Graph API integration (Office 365, Outlook, Teams)
• Multi-tenant SaaS architecture experience
• Familiarity with MSP/PSA ecosystems (ConnectWise, Datto, etc.)

Why Synthreo

• Work on AI that ships. Our platform is live, MSPs are paying, and your code will reach real users within weeks.
• High ownership, low bureaucracy. Small team. Direct founder access. Your decisions matter.
• Fully remote, async-first. We document everything so you can do deep work on your schedule.
• Competitive compensation. Salary tailored to your experience and location, plus equity options.
• Growth trajectory. We’re early. There’s room to take ownership, level up fast, and grow with the company.
• Learn from real AI systems. Work hands-on with LLM integrations, RAG pipelines, and agent orchestration — not toy projects.

How We Work

• Lean process: useful tests, clear contracts, production pragmatism over ceremony
• Async by default with synchronous overlap for collaboration when needed
• We ship incrementally — small PRs, fast feedback, continuous deployment
• Czechia-based candidates preferred, but we’re open to talent from anywhere

Our Commitment

Synthreo is committed to building a diverse team. We evaluate candidates based on skills and potential, not pedigree. We encourage applications from people of all backgrounds, including those who don’t check every box on this list.

How to Apply

Send your resume and a brief note about something you’ve built — a side project, a feature at work, or an AI experiment you’re proud of. We read every application and respond within one week.

[email protected]