We’re building AI-powered products, primarily leveraging off-the-shelf LLM APIs, while also partnering with our Applied Science teams to integrate custom models when needed. You’ll own end-to-end development of AI features—from prototyping to production—focused on reliability, accuracy, and scalable delivery.
Plan and develop applications that include custom AI components (orchestration, evaluation, tooling, and integrations).
Build intelligent agents using LLMs and the MCP framework to enable context-aware automation and decision-making.
Automate internal workflows and operational processes using Python and AWS Lambda.
Develop and improve RAG-based chatbots to increase response accuracy, grounding, and consistency.
Collaborate with Applied Science teams to productionize custom models and connect them with product systems.
Strong Python engineering skills and experience delivering production systems.
Hands-on experience building with LLM APIs (OpenAI, Anthropic, etc.) and deploying AI features end-to-end.
Experience building agentic workflows (tool use, routing, memory, safety/guardrails), ideally with MCP.
Solid understanding of RAG concepts (retrieval, chunking, embeddings, re-ranking, evaluation).
Practical AWS experience, especially AWS Lambda (plus IAM, CloudWatch, S3, API Gateway is a plus).
Comfortable working in fast-moving product environments with high ownership.
Experience integrating or serving custom models (inference, model endpoints, monitoring).
Familiarity with evaluation frameworks (offline tests, human feedback loops, regression suites).
Strong software engineering fundamentals (testing, CI/CD, observability, performance).
SEND YOUR RESUME IN ENGLISH, please.
Include:
Founded by James Sullivan, who built OneSeven Tech - a premier digital product agency serving startups and enterprises. OneSeven has partnered with startup clients who have collectively raised over $100M in VC, with enterprise clients including 2000+ person hospitality groups and NASDAQ companies.
Having experienced firsthand how tech debt can destroy startups, we're on a mission to turn the "invisible tax" of bad code into actionable business intelligence. We work remotely, move fast with 1-week sprints, and focus on elegant solutions to complex problems.