Forward Deployed Engineer - Architect

Jade Global•San Jose, CA

About The Position

The Forward Deployed Engineer - Architect leads customer discovery and translates customer priorities into practical solution designs and delivery plans. The role maps workflows, identifies bottlenecks, assesses data and system readiness, and establishes focused problems with measurable success criteria. It determines where AI, conventional software, or human review best fits the use case. The role builds production-quality AI applications, agents, and supporting services, including APIs, retrieval pipelines, tool integrations, orchestration logic, and user interfaces as needed. It integrates solutions with enterprise applications and data sources while addressing authentication, permissions, data quality, and reliable cross-system execution. The engineer establishes evaluation datasets and acceptance criteria, tests quality, task completion, failure modes, latency, and cost, and improves solutions before and after deployment. The position owns production readiness with customer and Jade teams, including access controls, auditability, monitoring, safe fallbacks, human approvals, operational handover, and post-launch troubleshooting. It provides technical direction while writing and reviewing code, manages dependencies and risks, communicates scope and architecture tradeoffs, measures adoption and business impact, and turns successful patterns into reusable components. The role also contributes field feedback to Jade Orbit and Jade's FDE delivery methods. Success is measured by time to a useful first release and progression to reliable production use; adoption and measurable workflow improvements; solution reliability, evaluation results, and operating-cost management; and customer trust, clear technical ownership, and reusable engineering assets. The role offers the opportunity to help shape Jade's growing AI Engineering Services business, work directly with enterprise customers, and contribute to its engineering practices, AI platforms, reusable solutions, and FDE-led delivery approach.

Requirements

At least 6 years of hands-on software engineering experience, including at least 2 years building LLM-powered applications or agentic systems.
Demonstrated ownership of multiple production deployments, with the ability to explain personal code contributions, architecture decisions, failure modes, and results.
Strong Python engineering skills, including asynchronous programming, API development, automated testing, and debugging.
Solid grounding in SQL, data structures, distributed systems, and application design, with the ability to contribute across backend and user-facing components.
Deep understanding of LLM application architecture, including model selection, embeddings, context management, structured outputs, tool calling, and inference tradeoffs.
Ability to balance model quality, latency, cost, and privacy requirements.
Hands-on experience engineering agents and orchestration using a framework or custom implementation.
Ability to implement explicit state, persisted checkpoints, retries, idempotent tool execution, human approvals, and recovery from partial failures.
Ability to build and tune retrieval-augmented generation pipelines, including document ingestion, chunking, embeddings, hybrid search, metadata filtering, reranking, and access-aware retrieval.
Experience diagnosing retrieval quality and grounding failures.
Experience developing rigorous AI evaluations with representative datasets, task-level metrics, regression tests, and adversarial cases.
Ability to measure factual accuracy, tool-use correctness, task completion, and failure rates rather than relying on subjective demonstrations.
Strong enterprise integration skills across APIs, authentication, databases, and event-driven systems.
Ability to expose business capabilities as secure tools and implement permission boundaries for agents that read data or execute actions.
Hands-on production deployment and operations experience on a major cloud platform, including containers, CI/CD, secrets management, tracing, monitoring, and incident diagnosis.
Ability to control concurrency, rate limits, inference costs, and operational reliability.
Practical experience securing AI applications against prompt injection, sensitive-data exposure, and unauthorized tool execution.
Ability to implement least-privilege access, audit trails, validation, and human review where consequences require it.
Ability to lead technical discovery and architecture discussions directly with customers, turn incomplete requirements into working software, and guide distributed engineers while continuing to write and review production code.

Nice To Haves

Previous work as an FDE, customer engineer, hands-on technical consultant, technical founder, or solutions architect with substantial coding and deployment ownership.
Experience with agent orchestration frameworks, Model Context Protocol integrations, LLM observability, evaluation pipelines, or model selection and fine-tuning when justified by the use case.
Experience building enterprise AI foundations, modernizing applications with AI, or working with open-source models and private deployment requirements.
Delivery experience in manufacturing, high-tech, or healthcare and life sciences.
Experience collaborating with distributed engineering teams.

Responsibilities

Lead discovery with business users and technology teams; map workflows, identify bottlenecks, assess data and system readiness, and agree on focused problems and measurable success criteria.
Translate customer priorities into solution designs and delivery plans, choosing appropriately among AI, conventional software, and human review.
Build production-quality AI applications, agents, and supporting services, including APIs, retrieval pipelines, integrations, orchestration, and user interfaces as required.
Connect solutions to enterprise applications and data sources, addressing authentication, permissions, data quality, and reliable execution across systems.
Create evaluation datasets and acceptance criteria; test task completion, output quality, failure modes, latency, and cost, then improve solutions before and after deployment.
Own production readiness with customer and Jade teams by implementing access controls, auditability, monitoring, safe fallbacks, human approvals, and operational handover.
Troubleshoot issues after launch.
Provide technical direction to the delivery team while writing and reviewing code.
Manage dependencies, surface risks early, and clearly explain scope, architecture, and delivery tradeoffs to customers.
Measure adoption and business impact against agreed baselines.
Turn successful patterns into reusable components and contribute field feedback to Jade Orbit and Jade's FDE delivery methods.