Senior Applied AI Engineer

This is a fully remote position, open to applicants in United States, +1 more state.

📋 Description

• Develop and launch Tango’s inaugural AI-driven product.

• Design, create, and deploy production-ready AI agents using LangGraph.

• Manage agents comprehensively, encompassing graph architecture, tool specifications, prompt and context engineering, sustained execution, failure and retry mechanisms, as well as cost and latency parameters.

• Construct agent tools interfacing with internal systems via MCP, utilizing direct API calls and agent-to-agent communication when suitable.

• Implement human-in-the-loop review processes with interruption points, confidence indicators, and correction pathways.

• Develop and optimize retrieval mechanisms addressing chunking, hybrid retrieval, grounding, and citations referencing source pages and paragraphs.

• Contribute to golden datasets, LLM-as-judge methodologies, deterministic scoring systems, and CI regression frameworks.

• Analyze quality issues and rectify them at the appropriate layer.

• Oversee agent-level safety protocols, including resistance to prompt injection, handling of PII, refusal processes, and escalation pathways.

• Collaborate with Product teams to convert accuracy benchmarks, confidence disclosures, and human-in-the-loop triggers into delivered functionalities.

• Partner with Platform Engineering for deployment and with Data Platform for curated datasets.

• Integrate agent session and usage analytics into the data warehouse.

• Transition reference agents to domain teams and aid in establishing a shared agent quality standard.

⛳️ Requirements

• Minimum of 7 years of professional software engineering experience.

• At least 2 years of experience building LLM-powered systems that have been deployed and utilized by real users.

• Strong proficiency in Python and its service ecosystem, including FastAPI, Pydantic, or similar technologies.

• Proven engineering practices in testing, code reviews, CI/CD, and production accountability.

• Experience in a production environment using an agent orchestration framework.

• Hands-on experience with at least one frontier model API.

• Practical knowledge of LLM evaluation, including golden datasets, LLM-as-judge, deterministic scoring, regression testing, and release management.

• Familiarity with MCP tool servers or similar tool and function-calling protocols.

• Experience with multi-agent design patterns.

• Background in production RAG and retrieval techniques, including chunking strategies, hybrid retrieval, grounding, citation, and diagnosing retrieval issues.

• Knowledge of LLM observability and tracing tools like LangSmith, Langfuse, Arize, or equivalent solutions.

• Experience in managing prompts and versioning.

• Sound judgment regarding hallucinations, prompt injection, silent degradation, and distinguishing acceptable from unacceptable failures.

• Preference for experience with LangGraph; familiarity with LangChain, OpenAI, or Claude Agents SDKs, or similar frameworks is also considered.

• Preferred background in graph-backed agent memory or knowledge graphs.

• Preferred experience in production vector and hybrid retrieval systems.

• Preferred experience in asynchronous task orchestration for extended document pipelines.

• Preferred experience in document intelligence and information extraction on a large scale, including OCR, layout-aware parsing, and structured extraction from lengthy documents.

• Applicants must possess authorization to work in the U.S. for any employer.

• Tango is unable to sponsor employment-based visas at this moment.

🏝️ Benefits

• Health, dental, and vision insurance.

• 401(k) plan with company matching.

• Generous paid time off policy.

• Fully remote work option.

• Inclusive and collaborative workplace culture.

• Competitive compensation package.