2 weeks ago

Senior Python AI Engineer (LLM & Multi-Agent Systems)

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI

Last checked: 21 hours agoStatus: Live

Job Summary

Build “Ask Seeking Alpha,” a high-load financial analysis system powered by Large Language Models. Design multi-agent orchestration with LangGraph, implement tool/function calling for accurate access to internal financial data, and optimize latency via async processing and streaming. Improve reliability of non-deterministic outputs using evaluation pipelines (LangSmith) and advanced RAG (hybrid search, re-ranking, query expansion). Work with OpenAI and AWS Bedrock models in production.

Location: Portugal

Workplace: Remote

Employment Type: Full time

Job Function: Data Science & Machine Learning

Seniority: Mid level

Key Responsibilities

•Design and implement multi-agent orchestration logic with LangGraph, including state management, conditional routing, and error handling.
•Build and optimize the tool layer (function calling) to let LLMs interact accurately with internal financial APIs and databases.
•Reduce end-to-end latency using asynchronous processing and streaming (SSE).
•Set up observability and automated evaluation pipelines with LangSmith, including regression testing for prompts and agents to measure quality before deployment.
•Refine retrieval strategies for advanced RAG using hybrid search (keyword + vector), re-ranking, and query expansion to provide relevant context to the model.

Key Requirements

•Strong modern Python skills, including asynchronous programming (asyncio) for a non-blocking I/O pipeline, plus experience with FastAPI and Pydantic (v2).
•Production experience with LangChain and hands-on (or deep conceptual) understanding of LangGraph (state-machine based agent frameworks).
•Experience managing LLM non-determinism, including strategies for hallucination handling and reliable outputs (e.g., self-correction loops, CoT/ReAct prompting).
•Ability to enforce structured outputs using schemas (Pydantic/JSON mode) for reliable downstream processing.
•Advanced context optimization techniques for limited context windows (e.g., summarization chains, sliding windows, selective context injection) and understanding inference economics (latency vs. cost).

Experience:FinTech

Tech Stack:PythonAsyncioFastAPIPydanticLangChainLangGraphElasticsearchAWS BedrockOpenAI APILangSmithSSESemantic cachingRAGHybrid searchRe-rankingQuery expansionJSON modeVector databasesEmbeddings

Company Brief

Seeking Alpha

Provides financial news, market analysis, and investment research for retail investors and professionals. The platform features articles, earnings coverage, stock ideas, portfolio tools, and community discussions focused on public markets.

Industry: News, Journalism & Broadcasting

Company Size: Large (251 to 1,000 employees)

Growth: Established Company

Headquarters: New York City, United States

Founded: 2004

Website LinkedIn

Loading job details...

Preparing the role view and application actions.

Seeking Alpha

2 weeks ago

Senior Python AI Engineer (LLM & Multi-Agent Systems)

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI

Last checked: 21 hours agoStatus: Live

Job Summary

Location: Portugal

Workplace: Remote

Employment Type: Full time

Job Function: Data Science & Machine Learning

Seniority: Mid level

Key Responsibilities

•Design and implement multi-agent orchestration logic with LangGraph, including state management, conditional routing, and error handling.
•Build and optimize the tool layer (function calling) to let LLMs interact accurately with internal financial APIs and databases.
•Reduce end-to-end latency using asynchronous processing and streaming (SSE).
•Set up observability and automated evaluation pipelines with LangSmith, including regression testing for prompts and agents to measure quality before deployment.
•Refine retrieval strategies for advanced RAG using hybrid search (keyword + vector), re-ranking, and query expansion to provide relevant context to the model.

Key Requirements

•Strong modern Python skills, including asynchronous programming (asyncio) for a non-blocking I/O pipeline, plus experience with FastAPI and Pydantic (v2).
•Production experience with LangChain and hands-on (or deep conceptual) understanding of LangGraph (state-machine based agent frameworks).
•Experience managing LLM non-determinism, including strategies for hallucination handling and reliable outputs (e.g., self-correction loops, CoT/ReAct prompting).
•Ability to enforce structured outputs using schemas (Pydantic/JSON mode) for reliable downstream processing.
•Advanced context optimization techniques for limited context windows (e.g., summarization chains, sliding windows, selective context injection) and understanding inference economics (latency vs. cost).

Experience:FinTech

Tech Stack:PythonAsyncioFastAPIPydanticLangChainLangGraphElasticsearchAWS BedrockOpenAI APILangSmithSSESemantic cachingRAGHybrid searchRe-rankingQuery expansionJSON modeVector databasesEmbeddings

Company Brief

Seeking Alpha

Industry: News, Journalism & Broadcasting

Company Size: Large (251 to 1,000 employees)

Growth: Established Company

Headquarters: New York City, United States

Founded: 2004

Website LinkedIn