AI Architect

Location: Remote (LATAM preferred, nearshore)
Engagement: Contract / Full-time
Experience: ~5-8 years in traditional terms, but we care more about AI-native architectural capabilities than tenure

About the Role

This is a product-minded, AI-native architectural role where you'll design and ship AI-first systems for real clients.

You won't be implementing tickets or following specs. You'll be architecting multi-agent systems, designing intelligent workflows, and making critical decisions about how AI integrates into production—while operating inside our unified AI-orchestration system that maintains full context across the entire SDLC.

What You'll Work On

Architecting AI-first systems across multiple domains:

• Consumer-facing intelligent applications

• Enterprise workflow automation with agentic orchestration

• Healthcare platforms and regulated industry AI solutions

• Internal operational AI tools and admin systems

Designing and implementing production-grade agentic workflows:

• Multi-agent system architecture and orchestration

• Context management strategies at scale

• Quality control, drift detection, and evaluation pipelines

• Cost optimization and token management

Owning AI architecture 0 → 1:

• System design for AI-native products

• LLM integration strategies and provider selection

• RAG pattern implementation and vector database design

• Agent debugging, monitoring, and optimization

Operating within Day10's AI-orchestration system:

• Leveraging our unified toolchain (V0, Claude Code, Cursor, proprietary agents)

• Maintaining architectural context across requirements, models, and systems

• Following AI-first SDLC methodology

Collaborating with product leadership and clients:

• Translating vague requirements into concrete AI architectures

• Unblocking technical decisions and moving fast

• Shipping intelligent systems, not experiments

Technical Requirements

Strong AI-native foundation:

• Backend: Python (FastAPI strongly preferred)

• AI Stack: Deep experience with LangChain, LangGraph, CrewAI, or similar frameworks

• LLM APIs: Hands-on with OpenAI, Anthropic, and open-source models

• ML Frameworks: PyTorch and/or TensorFlow (experience building, training and fine-tuning models)

• Vector Databases: Pinecone, Weaviate, Chroma, or similar

• Frontend: Familiarity with React (not mastery required, but you need to understand how UI integrates with AI systems)

AI system architecture ownership:

• You design multi-agent systems that work in production, not just prototypes

• You understand trade-offs between different agentic patterns

• You know how to manage context windows, retrieval strategies, and prompt chains

• You architect for cost, latency, and quality—not just functionality

Production AI experience:

• You've shipped AI systems where LLMs are core to the architecture

• You've debugged hallucinations, quality drift, and cost explosions

• You understand evaluation frameworks and how to measure AI system performance

• You've made architectural decisions under ambiguity with real production constraints

Strong AI systems design fundamentals:

• You think in agent workflows, context flows, and retrieval patterns

• You design evaluation pipelines and quality gates

• You consider cost optimization, caching strategies, and fallback mechanisms upfront

AI / Agentic Experience (Critical)

This is non-negotiable. This role is for people who live and breathe AI-first architecture.

Deep hands-on agentic development:

• You build with AI tools daily (Claude, Claude Code, Cursor, etc.)

• You've architected and shipped multi-agent systems in production

• You understand agent orchestration patterns: sequential, parallel, hierarchical, collaborative

• You've designed systems where agents call other agents and maintain state

Production AI architecture:

• You've designed RAG systems with vector databases and retrieval strategies

• You understand prompt engineering as an architectural discipline

• You've implemented quality monitoring, drift detection, and evaluation loops

• You know how to optimize for token usage and API costs without sacrificing quality

Debugging and optimization:

• You've debugged agent hallucinations and quality issues at scale

• You've optimized slow agentic workflows to meet production SLAs

• You understand how to trace and monitor multi-agent interactions

• You know when an agent-based approach is overkill vs. when it's necessary

What "Senior" Means at Day10 for AI Architects

We don't define seniority by years. We define it by architectural judgment + AI-native fluency.

A senior AI architect at Day10:

• Has designed and owned complex AI systems end-to-end in production

• Can architect, implement, and iterate independently on agentic workflows

• Knows when to ask clarifying questions—and when to make architectural calls themselves

• Is fluent in AI-augmented workflows and treats AI as the foundation, not an add-on

• Designs systems that are 3-4x more efficient than traditional architectures because they leverage AI intelligently

We hire AI-native architects who:

• Learn fast and evolve constantly with the AI landscape

• Obsess over system efficiency, cost, and quality

• Have the pragmatism to let AI handle orchestration

• Have the judgment to architect workflows that AI can reliably execute

What We Care About (Non-Negotiables)

Architectural ownership mentality

• You treat every system design like it's yours to maintain

• You don't get blocked—you make architectural decisions and move forward

• You ship intelligent systems, not just clever prototypes

Proactiveness

• You ask questions to clarify product goals, not to avoid technical decisions

• You dig for context when requirements are vague

• You identify architectural risks and quality issues before they hit production

Strong communication

• Clear, concise, and direct about technical trade-offs

• Active in Slack, engaged in architecture discussions

• Opinionated in a constructive way about AI patterns and approaches

Diligence & collaboration

• You follow through on architectural commitments

• You work well in small, high-trust, fast-moving teams

• You give and receive technical feedback openly

Smart architecture without over-engineering

• You design for production, not perfection

• You balance "ship now" vs. "architect right"

• You know when to refactor agent workflows and when to iterate

Nice to Have

• Experience in regulated domains (healthcare, fintech, insurance) where AI quality and explainability matter

• Product intuition—you care about why AI systems are built, not just how

• Previous startup or 0 → 1 AI product experience

• Exposure to MLOps/LLMOps practices and deployment pipelines

• Deep familiarity with agent evaluation frameworks and quality benchmarking (e.g., LangSmith, Weights & Biases)

• Experience with open-source LLMs and local model deployment