Mimir

Compensation

Full Description

Founding Engineer (Cloud & Full-Stack) Company: Bucket Labs, Inc. (Product: Mimir) Location: Remote (US-Only) or San Diego, CA Compensation: $80k – $160k Salary + 0.0% – 10.0% Equity Type: Full-time/Contract The Mission At Bucket Labs, we are building Mimir: the intelligent infrastructure for the next generation of private AI. We are moving beyond simple "wrappers" around hyperscaler APIs to build a secure, high-accuracy Retrieval-Augmented Generation (RAG) system that enterprises can trust with their most sensitive data with our next generation of Knowledge Management and Agentic AI. We are looking for a Founding Engineer who refuses to choose between "DevOps" and "Product." You will architect the cloud backbone that secures the data and write the code that makes it usable. If you want to own the stack from the React frontend down to the Terraform modules—and can explain every line of code in between—this is your arena. What You Will Build (The Hybrid Role) You will own the full lifecycle of our product ecosystem. You aren't just deploying; you are designing for resilience, cost-efficiency, and user experience. 1. The Product (Full-Stack RAG) Core RAG Experience: Build the user-facing application (Next.js/React) and the backend logic (Node/Python) that powers document upload, parsing, citation-based chat, and shared workspaces. Retrieval Engine: Implement and tune vector search (pgvector, Weaviate), chunking strategies, and embedding models. The Future Agents: Lay the groundwork for autonomous agents that will eventually interact with email, CRMs, and internal tools. 2. The Infrastructure (Cloud Architecture) Multi-Cloud Orchestration: Architect and deploy cloud-native systems across AWS (ECS, Lambda, Bedrock) and GCP (Cloud Run, GKE, Vertex AI). Infrastructure as Code (IaC): Maintain strictly reproducible environments using Terraform or CloudFormation. • Zero-Trust Security: *This is a private AI product. You will configure IAM, VPC networking, and KMS encryption to enforce airtight access control. Cost & Performance: You will aggressively optimize costs by choosing open-source components over expensive managed services where possible (e.g., self-hosting OCR vs. Textract). The Tech Stack We don’t expect you to know everything day one, but strong overlap here is required: Frontend: React, Next.js, TypeScript, Tailwind. Backend: Node.js and/or Python (FastAPI/Express). Cloud & Ops: AWS, GCP, Docker, Kubernetes/ECS, Terraform, GitHub Actions. Data: Postgres (pgvector/SQL), Redis, S3/Object Storage. AI/ML: LangChain/LlamaIndex, Open-source OCR (Tesseract), Vector DBs. Who You Are You are a Builder, not just a Maintainer. You have 4+ years of experience and have shipped significant features end-to-end, ideally in a startup environment. You value "Vibe Code." You write clean, testable code and can explain the purpose of every function and configuration. You are Cost-Conscious. You understand the trade-offs between "build vs. buy." You get excited about replacing expensive APIs with efficient open-source alternatives. You think in Systems. You understand how a React component interacts with an API gateway, how that hits the database, and how the VPC secures it all. You want Ownership. You are looking for a seat at the table with the founders, where your decisions on architecture and product direction matter immediately. Why Join Bucket Labs? We believe the future of AI is private, secure, and free from hallucinations. Founding Team Seat: You wil become a core member of the founding team with meaningful equity (up to 10%) and direct influence on the company's future. Bleeding Edge Tech: Work with the latest in RAG, Agentic AI, Vector Search, and local-first AI models. Freedom to Execute: No bureaucracy. If you can justify the tech choice, we will build it. Remote + HQ: Work from anywhere in the US, with the option to join us in sunny San Diego. Bonus: The "De-Clouding" & On-Prem Mindset We are actively moving toward self-sufficiency. While we currently use hyperscalers, our roadmap includes repatriating data and building our own compute infrastructure to maximize privacy and eliminate the "cloud tax." You are a massive plus if you bring: Bare Metal Experience: You’ve built systems outside of AWS/GCP (e.g., Colocation, dedicated servers, or heavy "Homelab" setups). Virtualization Mastery: Experience managing Proxmox, KVM, or VMware in production environments. Hardware Knowledge: You understand the physical side of servers, networking, and GPU clustering. The "Exit Strategy": You have ideas on how to migrate workloads from managed cloud services to self-hosted, open-source equivalents (e.g., self-hosting LLMs or Vector DBs) without sacrificing reliability. How to Apply Share your GitHub, LinkedIn, or Portfolio with the subject line: "Founding Engineer – I Build, Not Just Deploy" Please include a short note answering these two questions: Tell us about a RAG or search feature you owned end-to-end (stack used + what you’re proud of). Describe a time you avoided or replaced an expensive managed service (like Textract or a proprietary vector store) with an open-source or self-hosted alternative. What were the trade-offs?