### **About Us**
We are a **stealth-mode startup** building a new AI platform. Our mission is to make advanced language models deployable, customizable, and secure across diverse environments.
### **Role**
We are seeking a **Backend Engineer (Node.js/NestJS)** to extend our platform using our existing codebase. You’ll build the **proxy backend** that interacts with our custom inference runtime and extend dashboards.
This role requires strong backend engineering skills, the ability to integrate existing systems, and comfort working closely with C++/CUDA engineers building low-level runtime features.
### **Responsibilities**
#### **Proxy Backend for Inference Runtime**
– Build and maintain a **Node.js-based proxy backend**that:
– Accepts inference requests from the frontend.
– Schedules and serializes prompts.
– Manages **QKV cache load/unload**.
– Provides APIs to **manage LoRA adapters**.
– Integrate with authentication, RBAC, and logging already provided by the existing stack.
– Expose metrics and logs for monitoring inference usage and performance.
#### **Dashboards**
– Extend the existing Dashboard with Dataset upload, training job view, model management, inference usage, request history, and adapter selection.
– Reuse **auth, billing, and user management code** (Auth0, Stripe).
– Add necessary backend endpoints to support new UI flows.
#### **Core Stack & Infrastructure**
– Develop using **NestJS** as the main backend framework.
– Work with **PostgreSQL, Redis, and HashiCorp Vault** for persistence, caching, and secrets.
– Use **Socket.IO** for real-time updates (job status, inference progress).
– Ensure secure integration with Stripe (billing) and Auth0 (identity).
– Collaborate with DevOps on deployment pipelines (Proxmox, Docker, CI/CD).
### **Requirements**
– Strong experience with **Node.js** and **NestJS** framework.
– Proficiency in **PostgreSQL** and **Redis** for persistence and caching.
– Hands-on experience with **Socket.IO** or other WebSocket libraries.
– Experience with **secure configuration and secrets management** (HashiCorp Vault preferred).
– Comfortable working with **microservices** and integrating with existing codebases.
– Strong debugging and systems thinking — able to reason about scheduling, state management, and concurrency.
– Familiarity with LangChain for building AI-powered features and workflows.
### **Nice to Have**
– Experience integrating with **AI runtimes** (gRPC/REST backends for inference).
– Experience with RAG and MCP.
– Experience with **authentication/authorization frameworks** (Auth0, JWT, RBAC).
– Familiarity with **Stripe API** or similar billing systems.
– Contributions to backend open-source projects.
– Experience with WebRTC.
### **Why Join**
– Extend a proven SaaS foundation into a **new AI runtime platform**.
– Competitive compensation, equity potential.