You will be taken to the employer’s own careers page to apply.
Job details
Qualification
Master's / Bachelor's
Job ID
R0000397950
Job description
Career Area: Technology, Digital and Data Job Description: Your Work Shapes the World at Caterpillar Inc.
When you join Caterpillar, you're joining a global team who cares not just about the work we do – but also about each other. We are the makers, problem solvers, and future world builders who are creating stronger, more sustainable communities. We don't just talk about progress and innovation here – we make it happen, with our customers, where we work and live. Together, we are building a better world, so we can all enjoy living in it.
Job Description The Solutions Architect is the end-to-end technical owner of an enterprise analytics platform composed of a governed analytics function library, a visual workflow builder, an agentic AI analyst, and a tool/integration layer (e.g., Model Context Protocol servers) that grounds the agent in governed enterprise data. This role owns architectural cohesion, non-functional requirements (performance, scalability, reliability, security, cost), code quality, and delivery across all platform surfaces, and integration with upstream and downstream enterprise systems. As the platform's agentic capabilities mature, the role is the AI Solutions Architect — designing, reviewing, and governing agent, retrieval, evaluation, and safety architecture in line with enterprise architecture, security, and Responsible AI standards. What You Will Do Solution & Platform Architecture
Own the end-to-end architecture: data sources → analytics library → tool/integration layer → AI agent & workflow builder.
Define scalable, secure, reusable patterns for APIs, agent tools, event flows, and platform extensions; govern API lifecycle (versioning, backward compatibility, deprecation, contract testing).
Own non-functional requirements: latency budgets, throughput, availability/SLOs, capacity planning, and resilience (retries, circuit breakers, graceful degradation, DR/BCP).
Govern integrations with enterprise data sources (telemetry, service, warranty, parts, work orders) and ongoing data-warehouse migrations.
Maintain the architecture roadmap, reference architectures, and Architecture Decision Records (ADRs); drive build-vs-buy and vendor/tooling evaluations; manage tech debt deliberately.
Data, Retrieval & Context Architecture
Design the RAG / knowledge architecture: ingestion, chunking, embeddings, vector/hybrid search, retrieval quality, and freshness.
Own the semantic layer, data contracts, and metadata/catalog that let the agent reliably map natural language to governed data.
Define agent memory and context architecture (session state, short- vs long-term memory, context-window optimization, semantic caching).
AI & Agentic Architecture
Architect the agent backbone: orchestration/planner-executor patterns, tool boundaries, memory, evaluation hooks, and guardrails; support multi-agent patterns as needed.
Define the LLM provider strategy (cloud-hosted, self-hosted, model gateway) and a cost/FinOps architecture: model routing, caching, token economics, and budget guardrails.
Own the LLMOps lifecycle architecture: model/prompt versioning and registries, canary/A-B rollout of prompts, model upgrade and deprecation strategy.
Own the evaluation & observability architecture: LLM tracing (OpenTelemetry), trace stores, golden datasets, eval-in-CI, online/offline eval, drift and regression gates.
Establish AI governance practices: design reviews, risk assessments, Responsible AI alignment, and lifecycle management of agents and prompts.
Security, Trust & Compliance
Own the AI threat model: prompt injection, data exfiltration via tools, jailbreaks, output sanitization, and least-privilege tool authorization.
Architect multi-tenancy, data scoping, row-level security, PII/DLP, and audit/traceability across data and prompts.
Ensure alignment with enterprise architecture, IT controls, confidential-data handling, and emerging AI regulatory requirements (e.g., AI Act, model cards, data residency).
Agile Delivery & Technical Leadership
Lead sprint planning, code reviews, and technical design docs; set coding standards and release cadence.
Own end-to-end troubleshooting across all platform components and integration points.
Mentor engineers on architecture principles, integration patterns, and AI solution design; contribute to hiring, interviewing, team topology, and onboarding/enablement.
Stakeholder Collaboration & Governance
Serve as the trusted technical advisor to the PM, sponsor, AI Center of Excellence, enterprise IT, and security partners.
Lead or participate in architecture review boards, design-governance forums, and change-impact assessments.
Communicate complex architectural concepts and trade-offs to technical and business audiences.
What You Have
Education & Experience: Bachelor's or Master's in Computer Science, Engineering, or related field; 8+ years building production software, with at least 3 years as a tech lead or solutions architect on a data, analytics, or AI platform.
Application Design & Architecture: Converts business requirements into clean, modular, well-documented technical designs and reference architectures.
System & Technology Integration: Deep knowledge of integrating heterogeneous applications, databases, and platforms, with disciplined API/contract governance.
Systems & Reliability Engineering: Strong grounding in distributed systems, async patterns, scalability, observability, resilience, and secure-by-design.
Technical Troubleshooting: Anticipates, diagnoses, and resolves complex multi-system issues — including non-deterministic AI failure modes — quickly.
Technology Advising: Advises product, engineering, and senior leadership on technology trade-offs and risk.
Core Technical & Functional Skills
Languages & Frameworks: Expert-level Python; working knowledge of TypeScript / Node.js for full-stack reviews.
Cloud: Production experience on AWS and/or Azure (compute, storage, networking, IAM).
Data Platforms: Cloud data warehouse (e.g., Snowflake, BigQuery, Databricks); strong SQL; data modeling; vector stores and retrieval.
AI / Agentic: Hands-on with LLMs, agent frameworks (Model Context Protocol, function-calling, LangChain / LlamaIndex / Semantic Kernel / Agent SDKs), RAG, context/memory design, and evaluation harnesses.
