AI Engineering Lead

Engineering | Denver, CO or US Remote | Full-time

About Earlytrade

Earlytrade is a B2B fintech company built for the construction industry. We help general contractors and subcontractors improve cash flow through a platform that gives subcontractors more control over when they get paid and on what terms. The result is a fairer, more transparent way to manage working capital across the construction supply chain.

AI is not a side project at Earlytrade. We see it as a core lever for how the company scales: how subcontractors are reached, onboarded, activated, and supported; how internal teams create leverage; and how Earlytrade shows up inside the systems customers already use. This role is for someone who wants to build that capability from the ground up and turn AI into a real operating advantage, not a pile of demos.

The Opportunity

We’re hiring an AI Engineering Lead to be our first dedicated AI hire and to own how AI is rolled out across Earlytrade. This is a strategy-through-execution role: you’ll define where AI should create real advantage, build the platform and evaluation foundations that make that rollout safe and scalable, and ship production systems across outreach, onboarding, decision support, and embedded partner workflows.

You will work closely with the full engineering team plus product, revenue, and implementation stakeholders. This is a senior builder role first: we need someone hands-on enough to ship production systems now, with a clear trajectory toward building and leading an AI engineering function.

Success in year one is establishing the evaluation and governance foundations future AI work can safely build on, then shipping production systems that improve activation, create operating leverage, lower cost to serve, and open new ways for Earlytrade to appear inside the workflows our customers already live in.

What You’ll Build

• Earlytrade’s foundational AI platform: evaluation systems, prompt and workflow versioning, observability, governance, and cost controls that future rollout can safely build on
• Retrieval and context systems that let AI reason over Earlytrade’s operational knowledge, transaction history, and customer workflows with strong safeguards
• Applied AI products across subcontractor outreach, onboarding, decision support, and embedded partner experiences that create measurable commercial and operational leverage
• AI-assisted communication systems across email, SMS, and voice channels where automation creates real leverage without compromising quality or trust
• Decision-support systems that help teams route work, prioritize action, summarize context, and make better judgments faster in high-friction workflows
• Embedded AI experiences that allow Earlytrade functionality or intelligence to appear inside ERP, project-management, or partner workflows where that is strategically useful
• Model evaluation and benchmarking frameworks that measure accuracy, reliability, latency, cost, safety, and business impact for every AI system in production
• Prompt engineering systems, versioning infrastructure, and experimentation pipelines that support rapid iteration on model behavior without losing operational discipline
• Controlled fine-tuning and open-source model pathways where they create durable advantage, not just technical novelty
• Internal tooling that gives the broader Earlytrade team visibility into AI system performance, failure modes, governance, and cost attribution

Responsibilities

• Own the end-to-end AI rollout at Earlytrade: strategy, use-case selection, architecture, build, deployment, evaluation, and iteration
• Establish the AI platform standards that future rollout depends on: evaluation, regression testing, observability, prompt and workflow versioning, cost controls, fallback handling, and approval boundaries
• Design and ship production AI systems using provider APIs, orchestration frameworks, retrieval patterns, and supporting infrastructure where appropriate
• Build and maintain retrieval architectures including chunking strategies, embedding pipelines, vector stores, retrieval optimization, and context management
• Architect multi-step AI workflows and agentic systems with production-grade reliability, observability, and human-in-the-loop controls for high-risk actions
• Define and enforce evaluation standards: automated evals, human review workflows, regression testing, failure analysis, and business-impact measurement
• Partner with product, revenue, implementation, and engineering teams to translate real business bottlenecks into safe, measurable AI systems
• Select, evaluate, and integrate third-party AI infrastructure: model providers, vector databases, embedding services, voice AI platforms, communication APIs, and evaluation tooling
• Make sound buy-vs-build decisions across hosted models, open-source options, and future fine-tuning investments
• Define data-handling standards for AI systems, including PII safeguards, zero-retention API preferences where appropriate, human approval for sensitive communications, and controlled use of proprietary data for fine-tuning
• Maintain rigorous documentation of architecture decisions, evaluation results, data-handling assumptions, and known limitations
• Stay current with the rapidly evolving AI landscape, and make well-reasoned decisions about when to adopt new capabilities vs. when to stay disciplined
• Define the hiring roadmap and technical standards for Earlytrade’s AI function as the team scales

Required Qualifications

• 5+ years of professional software, ML, or applied AI engineering experience, with meaningful hands-on experience shipping LLM-powered systems in production
• A builder-first early hire: you move from ambiguous strategy to production systems without waiting for a large team or a perfect playbook
• Strong experience designing and shipping applied AI systems that solve real workflow or product problems, not just prototypes or notebooks
• Deep expertise in retrieval-based AI systems: embeddings, chunking strategies, vector stores, retrieval optimization, and context design
• Experience building multi-step AI workflows or agentic systems with strong fallback handling and human approval where risk demands it
• Strong judgment on when to use model APIs, retrieval, orchestration, workflow design, open-source models, or fine-tuning to solve a problem well
• Experience with evaluation frameworks: automated benchmarking, human eval pipelines, regression testing, failure analysis, and rollout measurement
• Strong engineering ability in Python, TypeScript, or similar production languages; you write clean, testable, well-documented code, not just experiments
• Experience working with major model provider APIs and understanding their capability, latency, cost, and data-handling tradeoffs
• Working knowledge of LLMOps or MLOps practices: deployment pipelines, versioning, monitoring, prompt and workflow iteration, and production alerting

Preferred Qualifications

• Experience fine-tuning or adapting models for domain-specific B2B applications, particularly in fintech, payments, or financial services
• Experience building AI-powered communication systems at scale: email, SMS, or voice, including personalization, deliverability, response handling, and compliance
• Familiarity with voice AI platforms and real-time speech-to-text pipelines
• Experience with parameter-efficient fine-tuning techniques: LoRA, QLoRA, adapter layers, or equivalent
• Familiarity with open-source model ecosystems and self-hosted or controlled deployment paths
• Experience with cloud AI or ML infrastructure on AWS or GCP
• Knowledge of data privacy, PII handling, and governance requirements relevant to AI systems processing financial and communications data
• Experience defining embedded AI experiences inside partner or third-party workflows is a strong plus
• Prior experience building or leading an AI function at a startup or early-stage company
• Exposure to the construction, real estate, or trade contractor industry is a genuine plus

Technical Context

• Rollout Posture: Platform foundations and evaluation discipline first, then applied AI rollout across outreach, onboarding, decision support, and embedded partner flows
• Current Product Stack: Java Spring Boot services, Node.js + TypeScript GraphQL, React + TypeScript frontends, PostgreSQL, and AWS
• Implementation Languages: Open-ended; Python is common for AI systems, but we care more about strong engineering judgment and the ability to integrate cleanly with the existing platform
• Core AI Capabilities In Scope: Model-provider APIs, retrieval systems, workflow orchestration, communication AI, observability, and tooling
• Data & Governance: PII-sensitive financial and communications data, strong safeguards, zero-retention preferences where needed, human approval on high-risk workflows, and controlled fine-tuning only
• Delivery Philosophy: Hybrid buy-and-build, principles first, open to hosted and open-source approaches where they materially improve speed, safety, performance, or economics

Why This Role

This is for someone who wants to define where AI belongs in a real business, build the foundations responsibly, and ship systems that change how Earlytrade reaches, activates, and serves customers.

The construction industry is years behind other verticals in AI adoption. The person who builds this foundation will have a meaningful head start, access to proprietary workflow and transaction context, and a direct line to the founders who are making the bets. The systems you build here can influence customer acquisition, operating leverage and embedded partner experiences, and the shape of the company’s next chapter.

What We Offer

• Meaningful equity: this hire is foundational
• Complete ownership of AI strategy and platform direction from day one
• Direct access to founders and a short decision-making chain
• Rich proprietary workflow and transaction context from real construction supply chain activity; a strong foundation for retrieval, evaluation, and controlled model improvement