AI Application Engineer Full Time NEW

Every solopreneur delivering a human-led service today is duct-taping five tools together just to get paid. Xuman.AI is replacing all of them. One AI-native operating system where content becomes discovery, discovery becomes booking, and booking becomes a relationship that compounds. We are live on both app stores with a real product, real architecture, and hard problems left to solve. If you are looking for a place where you can build fast, use AI as a force multiplier, and see your work in production within days – keep reading.

What you will work with:

– Python with FastAPI for production services

– Multi-model orchestration across leading LLM providers

– LangGraph for agent orchestration and multi-step workflows

– LiveKit for real-time voice AI

– WebSocket for bidirectional communication

– Deepgram and Azure Whisper for speech-to-text

– Docker and Azure Container Apps for deployment

What you will own:

– The AI layer of a live consumer product with real users

– A voice AI agent already deployed in production – you extend and improve it

– Safety and moderation systems for content and user protection

– Content generation pipelines for provider profiles and recommendations

– Business and user intelligence: analytics, personalization, behavior-driven features

– Prompt architecture, model selection, and AI reliability across all features

What we are looking for:

– 1-5 years of experience building Python production services (not scripting or notebooks)

– Hands-on experience with LLM APIs (OpenAI, Claude, or equivalent) and prompt engineering

– Understanding of real-time systems: WebSocket, streaming, event-driven architecture

– Experience designing system prompts, testing edge cases, and handling AI failures gracefully

– You use AI tools daily to accelerate everything – coding, debugging, research, system design. This is foundational to how we operate.

– Active builder – you experiment with new models, build AI tools, contribute to open source, or ship things outside of work

Nice to have:

– LangGraph or similar agent orchestration frameworks

– LiveKit or WebRTC experience

– Voice AI / STT / TTS pipeline experience

– FastAPI

Why join Xuman.AI:

You are joining the foundation. You will own the AI layer of a live product, with your work reaching users immediately. Equity discussion starts at 90 days based on impact – and at this stage, equity actually means something. The rare chance to build production AI systems from the ground up with full ownership.

How to apply:

We do not review resumes. We review how you think. Record a 5-7 minute video (YouTube unlisted or Google Drive) and email the link to Vishal@xuman.ai with subject “AI Engineer – [Your Name]”.

Part A (2 min):

Who are you? What is the most interesting thing you have built with AI/LLMs? Not what model you used – what problem you solved and how you thought about it. What did you learn about making AI reliable in production?

Part B (3-5 min): Download the app (App Store / Google Play) and explore xuman.ai. Then answer on camera:

(1) Where do you see the highest-impact AI opportunity in this product?

(2) If you were building a safety agent, what are the first 3 things it should detect?

(3) How do you test an AI feature before shipping?

Also include: a link to your GitHub profile or an AI project you have built.

• No video = no review. We review every submission personally within 48 hours.