# Solutions Architect IV - AI — Astreya

- Location: Remote, CA (remote)
- Posted pay: $114k–$180k/yr
- Posted: Nov 18, 2026
- Status: open (checked Nov 21, 2026)
- Apply (employer's site): https://astreya.wd5.myworkdayjobs.com/life-at-astreya-opportunities/job/Remote-CA/Solutions-Architect-IV---AI_R0018527
- Employer record: https://visajobs.io/company/astreya-partners.md

## Sponsorship history
- Tier: Good history
- Occupation (SOC): Software Developers (15-1252)
- Employer's certified LCAs for this occupation in CA, FY2024–FY2026: 1
- …in the whole US: 10
- With this exact title: 0
- Median DOL prevailing wage, level I / II: $135k / $162k
- Staffing / consulting firm: 25% of its H-1B applications place workers at client sites (FY2024–FY2026)

## Description
What this Job Entails:
We are seeking an experienced AI Solution Architect to design, build, and scale production-grade Generative AI and multi-agent platforms. Mirroring the delivery rigor required in regulated enterprise environments, this role bridges hands-on cloud AI engineering with consultative delivery leadership. You will architect shared, reusable AI platforms on AWS supporting high-stakes business domains—including regulatory compliance, energy contracts, finance, procurement, and operational quality—ensuring strict source grounding, cost efficiency, and measurable business return on investment (ROI).
Your Roles and Responsibilities:
Multi-Agent Platform & Cloud Architecture
Architect, deploy, and own production readiness for shared, multi-agent Generative AI platforms on AWS (Amazon Bedrock, AWS Lambda, Amazon ECS, Amazon OpenSearch, Terraform) that serve multiple business units from centralized enterprise knowledge bases.
Design multi-agent orchestration layers using frameworks such as LangGraph and LangChain , establishing reusable modular components and standardized design patterns that accelerate onboarding of new domain assistants.
Engineer advanced context-management strategies for long-running agent sessions—including on-demand skill and tool loading (progressive disclosure), Model Context Protocol (MCP) integrations, and prompt caching—to optimize latency and control token costs.
Enterprise Retrieval (RAG) & Ingestion Pipelines
Design and implement scalable document ingestion pipelines and hybrid retrieval architectures that combine semantic vector search, keyword search, and metadata filtering across large enterprise repositories (100+ GiB).
Establish chunking, reranking, and metadata strategies tailored to complex, heavily versioned utility documentation (such as regulatory filings, energy contract amendments, Standard Operating Procedures, and audit records).
Build strict guardrails for source grounding, citation traceability, and deterministic behavior within regulated quality and compliance workflows.
AI Evaluation, Observability & Governance
Develop rigorous, automated evaluation frameworks (including LLM-as-a-judge pipelines) that measure multi-dimensional answer quality, source recall and precision, and statistical confidence intervals across repeated runs.
Implement step-by-step agent observability and trace-based debugging using OpenTelemetry (OTel) and cloud evaluation tooling (such as AWS AgentCore Evaluations).
Design and execute structured User Acceptance Testing (UAT) protocols with business owners to validate traceability, accuracy, and compliance against business KPIs prior to production sign-off.
Technical & Delivery Leadership
Serve flexibly as Solution Architect, Tech Lead, or Delivery Lead depending on engagement needs—guiding cross-functional teams of data scientists and AI engineers through rapid prototyping, pilot validation, and full production rollout.
Partner directly with utility stakeholders across regulatory, legal/contracts, finance, and operations to translate complex workflows (such as tracing controlling language across contract amendments or automating regulatory filing quality checks) into high-ROI AI solutions.
Mentor junior engineers and developers on agentic design patterns, production software engineering best practices, and responsible AI governance.
Required Qualifications/Skills:
Required Qualifications
Experience: 8+ years of overall experience building enterprise data science, machine learning, and AI solutions, including 3+ years architecting and delivering production Generative AI and LLM applications.
Regulated Industry Delivery: Proven track record leading end-to-end AI delivery in regulated industries (Energy & Utilities strongly preferred; Financial Services, Telecommunications, Manufacturing, or Healthcare also relevant).
Production Impact: Demonstrated experience taking multi-agent or RAG platforms from prototype to active production adoption with measurable business impact (such as documented hours saved, cost reduction, or operational efficiency gains).
Cloud & Infrastructure (AWS): Hands-on architecture and deployment experience with Amazon Bedrock, OpenSearch, AWS Lambda, ECS, serverless and containerized AI patterns, cross-account environment promotion, and Infrastructure as Code (Terraform).
Agentic AI & Orchestration: Deep expertise in LangGraph, LangChain, prompt and context engineering, tool design, multi-agent workflows, prompt caching, and Model Context Protocol (MCP).
RAG & Evaluation: Practical mastery of hybrid search, metadata filtering, reranking, LLM-as-a-judge evaluation suites, grounding and citation verification, and OpenTelemetry trace debugging.
Programming & Data Engineering: Strong proficiency in Python, SQL, and FastAPI, with experience building scalable data and document ingestion and cleansing pipelines.
Education: Master’s or Bachelor’s degree in Information Systems, Data Science, Computer Science, Business Analytics, or a related quantitative discipline.
Preferred Qualifications
Prior technology or management consulting experience leading client-facing technical delivery, executive demonstrations, and UAT cycles.
Direct experience building AI assistants for energy contracts, regulatory filings, procurement, or quality and compliance workflows.
Familiarity with model fine-tuning and transfer learning, fairness metrics and bias mitigation frameworks (such as AIF360), and BI visualization tools (such as Power BI).
Relevant cloud or AI architecture certifications (such as AWS Certified Solutions Architect, AWS Certified Machine Learning, or Anthropic Certified Architect).
Physical Demand & Work Environment:
Must have the ability to perform office-related tasks which may include prolonged sitting or standing
Must have the ability to move from place to place within an office environment
Must be able to use a computer
Must have the ability to communicate effectively
Some positions may require o

> History, not a promise: past H-1B filings do not mean an employer will sponsor you or a given role. Data: U.S. DOL OFLC LCA disclosure (FY2024–FY2026 for role counts, certified LCAs) and the USCIS H-1B Employer Data Hub. Not legal advice. Methodology: https://visajobs.io/methodology
