Agentic AI Engineer

Fervo Energy; Houston (TX); Hybrid; USD 103,152 - 158,588; Full time

Benefits offered by this job

Comprehensive benefits package

Flexible paid time off

401(k) plan with employer match

Job summary

Fervo Energy in Houston, TX, is seeking an Agentic AI Engineer to design and deliver AI workflows that improve geothermal power operations. The role involves collaborating with various teams for successful production deployments and continuous improvement.

Candidates should possess strong Python skills, experience with AI applications, and familiarity with cloud environments. This position offers hybrid work options and a comprehensive benefits package.

Qualifications

• 2+ years of hands-on experience building agentic AI applications.
• Experience implementing architectural frameworks.
• Strong Git and CI/CD experience.

Responsibilities

• Design end-to-end agentic AI workflows.
• Build retrieval pipelines over technical documents.
• Deploy agents as production services.

Skills

Strong Python skills

Experience with agent frameworks

Knowledge of LLM APIs

Cloud deployment experience

Education

Bachelor's or Master's degree in Computer Engineering or Data Science

Tools

Azure

Docker

Terraform

Job description

Job Type

Full-time

Description

Fervo is building the most cost-effective, repeatable geothermal power plants in the world. Scaling that mission requires AI-native capabilities that drive measurable impact across drilling, completions, production, geophysics, and power plant operations. The Agentic AI Engineer, within the Data & AI team, designs and ships agentic workflows that turn unstructured knowledge and structured operational data into autonomous capabilities for engineers, operators, and decision‑makers in the field.

The Agentic AI Engineer owns the end-to-end delivery of agentic AI use cases — from problem framing and architecture, through prototyping, evaluation, deployment, and iteration in production. Working across Data Engineering, IT Infrastructure, domain SMEs, and business stakeholders, this role establishes reusable patterns for retrieval, semantic grounding, tool integration, and agent orchestration on top of our Azure, Databricks, and Snowflake stack. Success requires strong hands‑on engineering depth, sound architectural judgment, and pragmatism about what to ship versus what to defer.

Responsibilities

Agentic Workflow Design & Delivery

• Design and deploy end-to-end agentic AI workflows using planner–worker, orchestrator–executor, multi‑agent, and RAG‑based architectures
• Build reusable components and reference patterns for tool routing, state management, error handling, and human‑in‑the‑loop checkpoints
• Implement robust retrieval pipelines (hybrid search, vector + keyword, graph‑aware retrieval) over technical documents, historian data, and operational records
• Translate domain problems from drilling, completions, production, geophysics, and power plant operations into well‑scoped agentic use cases with clear success metrics

Semantic Grounding & Knowledge Integration

• Build and maintain a semantic layer over our data lake and warehouse using Snowflake Semantic Views and Databricks Unity Catalog Metric Views, making business concepts queryable by both humans and agents
• Develop and curate knowledge graphs that connect domain entities (wells, pads, assets, equipment, events, documents) and serve as grounding context for LLM reasoning
• Standardize how agents access enterprise data through Model Context Protocol (MCP) servers and equivalent integration patterns

Evaluation, Observability & Production Operations

• Establish agent evaluation frameworks including golden datasets, automated regression tests, and structured evals for accuracy, faithfulness, and tool‑use correctness
• Implement tracing, logging, and observability across agent runs to support debugging, cost monitoring, and continuous improvement
• Build feedback loops that capture user input and convert it into eval cases and prompt/system improvements
• Support production incidents and platform‑level issues impacting deployed agents

Deployment & Enablement

• Deploy agents as production services on our Azure‑native stack (App Service, Container Apps, Functions) with Entra ID SSO, Key Vault‑managed secrets, and proper cost controls
• Build lightweight UIs (Streamlit, Gradio, or React) for agentic applications and internal tools
• Lead design reviews and cross‑functional enablement sessions on agentic AI patterns and best practices

Qualifications

Required

• Bachelor's or Master's degree in Computer Engineering or Data Science preferred.
• 2+ years of hands‑on experience building and deploying agentic AI or LLM‑powered applications in production, not just prototypes or notebooks
• Strong Python skills, including async patterns, API design with FastAPI, and writing testable, maintainable production code
• Demonstrated experience with at least one major agent framework: LangChain/LangGraph, LlamaIndex, AutoGen, or Semantic Kernel
• Working knowledge of LLM APIs and SDKs (Anthropic Claude, OpenAI, Azure OpenAI), including tool use/function calling, structured outputs, streaming, and prompt engineering
• Experience implementing RAG architectures with vector databases (Azure AI Search, pgvector, Pinecone, Weaviate, Chroma, or similar) and embedding models
• Experience with agent orchestration patterns including multi‑step planning, tool routing, state management, and graceful failure handling
• Familiarity with Model Context Protocol (MCP) or equivalent standards for tool and context integration
• Cloud deployment experience on Azure (App Service, Container Apps, Functions, Key Vault, Entra ID), or equivalent in AWS/GCP with willingness to work in our Azure‑first environment
• Strong Git and CI/CD experience, including version control discipline, code review, and automated testing
• Experience with containerization (Docker) and infrastructure‑as‑code (Terraform preferred)
• Strong observability and production operations skills, including structured logging, tracing, cost monitoring, and runbook development

Preferred

• Experience designing or working with semantic models and semantic layers — Snowflake Semantic Views, Databricks Metric Views, dbt Semantic Layer, Cube, or Power BI semantic models
• Hands‑on experience with knowledge graphs: graph databases (Neo4j, Azure Cosmos DB Gremlin), RDF/SPARQL, ontology design, or graph‑augmented RAG
• Experience with agent evaluation tooling such as LangSmith
• Experience with our broader data stack: Databricks, Snowflake, Azure Data Lake Storage (ADLS), Azure Data Factory
• Oil and gas or energy industry experience, including familiarity with drilling, completions, production, geophysics, or industrial historian data
• Background in time‑series data, signal processing, or industrial IoT (MQTT, OPC UA, SparkplugB)

Experience with multimodal models for handling well logs, schematics, or scanned reports

Location

Fervo Energy is headquartered in Houston, TX, with growing offices in Golden, CO, Reno, NV, and Oakland, CA, and Salt Lake City, UT. This position will be eligible for some hybrid work flexibility, but regular in‑office presence at the Golden or Houston office will be required.

Compensation & Benefits

Fervo provides a comprehensive suite of benefits including medical, dental, vision, life, short‑term and long‑term disability, flexible paid time off, and paid parental leave. Additionally, Fervo offers an incentive stock options program, a bonus incentive program, and a 401(k) plan with an employer match.

Fervo Energy is providing the compensation range and general description of other compensation and benefits that the company in good faith believes it might pay and/or offer for this position based on the successful applicant’s education, experience, knowledge, skills, and abilities in addition to internal equity and geographic location. Expected Salary: $103,152 - $158,588 based on location and experience.

Fervo Energy reserves the right to ultimately pay more or less than the posted range and offer other compensation, depending on circumstances not related to an applicant’s sex or other status protected by local, state, or federal law.