Lead Developer, BlueSteel AI Solutions

Architecture and technical decisions

You decide how a build should be structured. This comes with a lot of ownership over every engineering project you are assigned to.

Frontend and backend implementation

Making decisions such as hosting on cloud platforms or using self-managed infrastructure, what programming language to use for a project, what APIs and libraries to use, how to structure database schema, etc.

AI integration as a core engineering discipline

Designing model chains, owning prompt quality and cost, and shipping features that turn LLM capabilities into reliable user-facing systems.

Production operations

Deployment, reliability, debugging, and ongoing maintenance of systems you ship.

Code review and technical direction for junior developers

A critical component of this role is managing a team of other developers. This primarily looks like delegating tasks to other developers and reviewing their work before committing it to deployment.

Technical interface with clients

You will participate in meetings and answer technical questions from clients.

Required

• 2+ years building production full-stack applications with React, Node.js, CSS (Tailwind), backend databases like MongoDB or Supabase, and Vercel or other frontend hosting services.
• Experience with systems programming languages like C or similar, unix/linux, and high level languages like Python.
• Hands-on experience deploying and operating services on VMs with Docker, including networking via tools like Cloudflare tunnels, nginx, or similar.
• Production experience integrating LLM APIs (Anthropic, OpenAI, or comparable) into real applications, including prompt design, cost control, and prompt caching.
• Comfort and experience owning a system end-to-end. Experience with continuous deployment and integration, debugging cloud services, managing API keys.
• Clear written communication. Most of our team coordination happens async in Slack with daily updates.

Strongly preferred

• RAG implementation experience in production, ideally with vector stores, hybrid search, or graph-based retrieval (GraphRAG, Microsoft GraphRAG, or comparable).
• Familiarity with Anthropic's Claude API specifically, including tool use, prompt caching, and model selection tradeoffs across Haiku, Sonnet, and Opus.
• Experience building workflow automation systems.
• Comfort working in a small team where you wear multiple hats (i.e. experience working in the startup space).

Not required, but a plus

• Open-source contributions, especially in the AI tooling, RAG, or developer-tools space.
• Experience self-hosting LLMs.
• Background working with non-technical stakeholders (founders, editors, subject-matter experts) and translating their needs into engineering solutions.

Tooling

GitHub for code. Slack for team and client comms. Claude Teams as the primary AI workspace.

Cadence

Weekly client standups, daily async updates from the dev team, compiled weekly status to client side. No daily standup meetings.

Ownership

You are trusted to make architectural calls and defend them. As with any management/lead positions, you are accountable for the success of projects and the efficiency of your team.

Confidentiality

Secure management of confidential client info and data is critical. Contractor terms are confidential to leadership.