Applied AI Engineer

Summary

Nexera is an AI-native consulting firm that builds intelligent, custom solutions for organizations navigating the age of AI. The Applied AI Engineer will build production AI applications and knowledge layers for secure, self-hosted platforms, including enterprise search, retrieval-augmented generation, agents, and developer tools. The role also involves evaluating, operating, and deploying AI systems while partnering with clients and engineering teams.

Responsibilities

• Design and Build Applied AI Solutions: Create production AI applications that combine language models, enterprise data, retrieval, agents, tools, and user-facing workflows to solve real business and engineering problems
• Connect Enterprise Knowledge: Build reliable ingestion and integration pipelines that make documents, repositories, applications, and organizational data available to AI systems under the right access controls
• Engineer Retrieval and Grounding: Develop and improve search, RAG, reranking, context assembly, citation, provenance, and authorization patterns that produce relevant, traceable, trustworthy results
• Create User and Developer Experiences: Build chat, research, workflow, API, IDE, and developer-tool experiences that make advanced AI capabilities feel obvious in everyday work
• Evaluate and Operate AI Systems: Establish tests, evaluation frameworks, observability, security controls, feedback loops, and release practices that raise quality and reliability over time
• Partner and Transfer Knowledge: Work directly with clients, users, subject-matter experts, and engineering teams to define needs, communicate tradeoffs, ship working solutions, and enable long-term ownership
• Innovate How We Work: Experiment with and define the agentic workflows and tooling that improve how Nexera designs, builds, and evaluates AI applications. Your patterns become part of our operating model

Skills

• Engineering experience that shows. Production software, data systems, enterprise search, developer tools, or AI-enabled products. Roughly three to five years, or equivalent capability you can demonstrate
• Hands-on applied AI work. You have built LLM applications, RAG systems, search experiences, agents, or local AI integrations in production
• Production software fundamentals. You can design maintainable services, data models, APIs, background workflows, tests, deployments, monitoring, and failure handling
• Full-stack capability. Strong in Python or an equivalent backend language, and able to build or meaningfully contribute to a modern TypeScript/React experience
• Retrieval judgment. Embeddings, lexical and vector search, hybrid retrieval, metadata filtering, reranking, chunking, and context construction, with an understanding of how to drive and improve both retrieval quality and answer quality
• Data-pipeline discipline. You have worked with ingestion, ETL, synchronization, retries, backpressure, deduplication, schema evolution, revisions, deletions, and operational observability
• Evaluation mindset. You can define representative test sets, pick metrics that mean something, run real error analysis, and build release gates that catch regressions before users do
• Security and authorization awareness. You have a deep understanding of application security and you can reason about identity, ACLs, least privilege, data lifecycle, prompt injection, provenance, audit, and cross-user leakage
• Product judgment. You care how streaming, sources, tool activity, errors, long-running work, and recovery feel to a user. You can turn complicated AI behavior into an experience people trust
• Daily AI tool user. You already use AI tools (Claude, Claude Code, Copilot, Cursor, or similar) in how you code, research, and solve problems, and you can clearly show and demonstrate how you apply them to day-to-day engineering work
• Relentless curiosity. You follow new model and tool releases because you find it interesting. You experiment on your own time. You have opinions about where this is going
• Strong communication skills. You can work with users, subject-matter experts, client engineers, data owners, security teams, and program leaders to turn ambiguous needs into tested, supportable capabilities
• Consulting mindset. You take ownership, communicate progress and risk early, document decisions, and adapt to the constraints of a client's environment
• U.S. citizenship and residency: Required for this role due to the customer environments and information the work may support
• U.S. citizenship and current U.S. residency are required
• Must be willing and able to complete applicable customer background investigations, satisfy site-access and data-handling requirements, and obtain and maintain a U.S. government security clearance if required for assigned work
• Expect approximately 25% travel, concentrated around discovery, integration, user testing, deployment, and major project milestones, with lighter travel between them. On-site work is scheduled in advance wherever possible
• Prior U.S. military service, experience supporting defense or national-security organizations, and/or a current or previously held U.S. government security clearance are preferred but not required
• Backend and data: Python, FastAPI, Django or Flask, asynchronous workers and queues, PostgreSQL, SQL, Redis, REST APIs, SSE or WebSockets
• Frontend and chat: TypeScript, React, Next.js or comparable frameworks, streaming chat interfaces, Markdown/code rendering, accessibility, responsive product development
• Retrieval and search: pgvector, PostgreSQL full-text search, Elasticsearch/OpenSearch, Solr, Vespa, Qdrant, Weaviate, Milvus, FAISS, HNSW, IVFFlat, BM25, hybrid retrieval, rerankers
• Document processing: IBM Docling, Apache Tika, Unstructured, OCR tools, PDF and Office parsing, scanned-document and multimodal pipelines
• File services and identity: SMB/CIFS, NFS, Active Directory, POSIX or Windows ACLs, LDAP, FreeIPA, Keycloak, OIDC, JWT, row-level security
• LLM applications and agents: LangGraph, LlamaIndex, LangChain, custom workflow engines, MCP clients and servers, structured output, tool calling, durable state, human approval and sandboxed execution
• Evaluation and observability: Ragas, DeepEval, promptfoo, Phoenix, Langfuse, OpenTelemetry, Prometheus, Grafana, structured logging, golden-set and regression harnesses
• User and research experiences: Open WebUI, LibreChat, Open Deep Research or comparable systems, citation and provenance interfaces, long-running research workflows
• AI CLI and agentic tools: Claude Code, Gemini CLI, OpenAI Codex, GitHub Copilot
• Developer experience: VSCodium or VS Code extensions, Cline, Continue, Tabby, aider, OpenHands, Goose
• Additional differentiators: knowledge graphs, GraphRAG, Apache AGE or Neo4j, bi-temporal data models, multimodal retrieval, regulated or classified environments, and disconnected or air-gapped deployments

Qualifications

Must Haves

• Engineering experience that shows. Production software, data systems, enterprise search, developer tools, or AI-enabled products. Roughly three to five years, or equivalent capability you can demonstrate
• Hands-on applied AI work. You have built LLM applications, RAG systems, search experiences, agents, or local AI integrations in production
• Production software fundamentals. You can design maintainable services, data models, APIs, background workflows, tests, deployments, monitoring, and failure handling
• Full-stack capability. Strong in Python or an equivalent backend language, and able to build or meaningfully contribute to a modern TypeScript/React experience
• Retrieval judgment. Embeddings, lexical and vector search, hybrid retrieval, metadata filtering, reranking, chunking, and context construction, with an understanding of how to drive and improve both retrieval quality and answer quality
• Data-pipeline discipline. You have worked with ingestion, ETL, synchronization, retries, backpressure, deduplication, schema evolution, revisions, deletions, and operational observability
• Evaluation mindset. You can define representative test sets, pick metrics that mean something, run real error analysis, and build release gates that catch regressions before users do
• Security and authorization awareness. You have a deep understanding of application security and you can reason about identity, ACLs, least privilege, data lifecycle, prompt injection, provenance, audit, and cross-user leakage
• Product judgment. You care how streaming, sources, tool activity, errors, long-running work, and recovery feel to a user. You can turn complicated AI behavior into an experience people trust
• Daily AI tool user. You already use AI tools (Claude, Claude Code, Copilot, Cursor, or similar) in how you code, research, and solve problems, and you can clearly show and demonstrate how you apply them to day-to-day engineering work
• Relentless curiosity. You follow new model and tool releases because you find it interesting. You experiment on your own time. You have opinions about where this is going
• Strong communication skills. You can work with users, subject-matter experts, client engineers, data owners, security teams, and program leaders to turn ambiguous needs into tested, supportable capabilities
• Consulting mindset. You take ownership, communicate progress and risk early, document decisions, and adapt to the constraints of a client's environment
• U.S. citizenship and residency: Required for this role due to the customer environments and information the work may support
• U.S. citizenship and current U.S. residency are required
• Must be willing and able to complete applicable customer background investigations, satisfy site-access and data-handling requirements, and obtain and maintain a U.S. government security clearance if required for assigned work
• Expect approximately 25% travel, concentrated around discovery, integration, user testing, deployment, and major project milestones, with lighter travel between them. On-site work is scheduled in advance wherever possible

Nice to Haves

• Prior U.S. military service, experience supporting defense or national-security organizations, and/or a current or previously held U.S. government security clearance are preferred but not required
• Backend and data: Python, FastAPI, Django or Flask, asynchronous workers and queues, PostgreSQL, SQL, Redis, REST APIs, SSE or WebSockets
• Frontend and chat: TypeScript, React, Next.js or comparable frameworks, streaming chat interfaces, Markdown/code rendering, accessibility, responsive product development
• Retrieval and search: pgvector, PostgreSQL full-text search, Elasticsearch/OpenSearch, Solr, Vespa, Qdrant, Weaviate, Milvus, FAISS, HNSW, IVFFlat, BM25, hybrid retrieval, rerankers
• Document processing: IBM Docling, Apache Tika, Unstructured, OCR tools, PDF and Office parsing, scanned-document and multimodal pipelines
• File services and identity: SMB/CIFS, NFS, Active Directory, POSIX or Windows ACLs, LDAP, FreeIPA, Keycloak, OIDC, JWT, row-level security
• LLM applications and agents: LangGraph, LlamaIndex, LangChain, custom workflow engines, MCP clients and servers, structured output, tool calling, durable state, human approval and sandboxed execution
• Evaluation and observability: Ragas, DeepEval, promptfoo, Phoenix, Langfuse, OpenTelemetry, Prometheus, Grafana, structured logging, golden-set and regression harnesses
• User and research experiences: Open WebUI, LibreChat, Open Deep Research or comparable systems, citation and provenance interfaces, long-running research workflows
• AI CLI and agentic tools: Claude Code, Gemini CLI, OpenAI Codex, GitHub Copilot
• Developer experience: VSCodium or VS Code extensions, Cline, Continue, Tabby, aider, OpenHands, Goose
• Additional differentiators: knowledge graphs, GraphRAG, Apache AGE or Neo4j, bi-temporal data models, multimodal retrieval, regulated or classified environments, and disconnected or air-gapped deployments

Benefits

• Performance bonus for full-time employees
• Flexible PTO plus federal holidays
• Annual learning and certification budget
• Access to premium AI tooling (Claude Code and the frontier tools you will use daily)
• Home-office setup
• Sensible travel policy — you book what makes the trip productive, and travel time is respected as work time
• Remote-first (US-based), with scheduled on-site work at client sites