Forward Deployed AI Engineer (Senior)

**_About AZX_** Our mission is to accelerate positive impact in critical industries through AI transformation. We’re growing quickly and already work with category-leaders in real estate (CBRE), energy (LevelTen Energy), logistics (Flexe) and utilities. We’re a public benefit corporation, founded in 2024 and have been profitable from inception. We work on challenges in clean energy, decarbonization, climate risk, energy systems and global economics. We’re building our company for long term success and aim to build the ultimate place to work if you’re passionate about AI and positive impact. **_About the Role_** We are seeking a Forward Deployed Engineer to join a team that scopes, builds, deploys, and measures AI systems inside client environments — utilities, commercial real estate, and logistics. Your software runs in the client's cloud, under their identity provider and toolchain, inside their compliance framework, integrated with their systems of record — you bring the right mix of models and tools for the job rather than working around it. The work is agentic AI with a correctness envelope: think document intelligence with deterministic, auditable validation where money or compliance is on the line, voice-of-customer AI, cognitive digital twins of a client's customers, and human-in-the-loop agentic workflows, all measured against real before/after baselines rather than left as shelf-ware. You won't be handed a finished spec — you'll sit with the client's operators and executives to find the real problem, design the solution with their architects, build it with your pod, and prove it worked with numbers both companies stand behind. You bring a strong area of expertise and expect to use it, and work closely with our team across DevOps, infrastructure, data pipelines, front end, and back end as the engagement requires — you are the engineering face of AZX. **_Responsibilities:_** - Own your client's technical delivery end to end — discovery support, solution design, build, deployment into the client's environment, and a handover their team can actually run. - Build the trust machinery behind every system you ship: eval harnesses, replay loops, guardrails, cost/latency budgets, monitoring, and a defined "what happens when it's unsure" path. - Extract structured facts from messy documents (like contractor bids or engineering forms) and build the deterministic checks that gate money- or compliance-sensitive answers, making "cannot determine" fail closed rather than open. - Optimize pipelines for cost and quality — for example, moving a step from a frontier model to a fine-tuned small model or a deterministic rule, then proving quality held with a replay harness. - Design and ship high-stakes systems like after-hours voicemail triage, defining the right autonomy boundary for the risk involved. - Agree on and track the measurement story with the client — KPIs, baselines, and instrumentation for cost, performance, and quality — in writing before deployment and validated after. - Maintain client-facing engineering presence: working sessions with their IT/security teams, demos, POCs that derisk the next engagement, and a feedback loop that carries field-learned requirements back to the platform team. **_Core Qualifications:_** - 5+ years of shipping LLM/agentic systems to production users — not prototypes — with structured outputs, tool use, retrieval, guardrails, and an eval loop you can defend to a skeptic. - A well-stocked technical toolkit and the judgment to use it: small task models (OCR, ASR, classification, reranking), classical NLP, fine-tuning/distillation, deterministic rules, caching, and a frontier model only where it earns its cost. - Full-stack delivery skills: Python/FastAPI backends, React/TypeScript front ends, deployment, monitoring, real test coverage, and careful data handling — since your pod is the whole team, there's no "someone else's layer." - Experience deploying inside someone else's cloud, identity provider, repos, and compliance regime, and negotiating their IT constraints without losing the design. - Strong stakeholder skills — running discovery with front-line operators, delivering executive readouts, and pushing back plainly (with a cheaper or safer alternative already sketched) when the ask is wrong. - Comfort with ownership under ambiguity — given a vague problem and a deadline, you return with a working thing or a good question - Practical fluency across our stack — Python (async/FastAPI/Pydantic), TypeScript/React, Postgres/pgvector, Redis, LLM provider APIs, RAG/hybrid retrieval, and agent frameworks (LangGraph/AutoGen/CrewAI-class or hand-rolled loops). - Familiarity with enterprise deployment concerns: Docker, Terraform/Bicep, Azure and/or AWS, enterprise SSO (SAML/OIDC, Entra), and observability/cost tracking. - A track record with enterprise integrations (SharePoint, Salesforce, SAP/ERP-class systems) - Domain exposure to utilities, commercial real estate, or logistics is a plus - Bachelor's Degree; Master's is a plus **_Why AZX!_** - Be part of a fast-growing, profitable, mission-driven company with industry-leading clients tackling the massive opportunity of AI transformation in critical industries. - Competitive early-stage startup compensation (based on capabilities, experience, and location) - Bonus eligibility - Health insurance with meaningful coverage for dependents - Flexible paid time off - Equity - Fully remote culture with a cluster of teammates in Seattle **_Additional Information:_** - Must be willing to travel to Seattle area for final interview and travel 2x/year for company summits - **Applicants must be currently authorized to work in the United States on a full-time basis.** - We are unable to sponsor or take over sponsorship of employment visas at this time. - Please note that our interview process includes a **written take-home assignment followed by a live two-hour technical session** with our engineering team, so if that format isn't a good fit, we'd ask that you not apply **_Next Steps:_** If this job sounds like a great fit but don’t check **ALL** of these qualification boxes, we’d still love to hear from you!