Principal Applied AI Architect

Order.co is the System of Action for the Office of the CFO, transforming the way businesses purchase and pay into an intuitive, B2C-like shopping experience. Order.co leverages embedded AI agents and embedded financial products to reinvent the way businesses connect with their vendors.

End users enjoy a seamless, zero-training buying experience, while finance and procurement leaders gain a single platform to orchestrate how the business “should operate”. The result is an all-in-one solution that serves as a gravitational pull for spend and data, automating and eliminating procurement and finance workflows from requisition to reconciliation along the way.

Order.co is on the cutting edge of B2B Agentic Commerce, poised to be the market leader in creating a more predictive, prescriptive, and personalized experience for users.

Founded in 2016 and headquartered in New York City, Order.co oversees nearly half a billion in annualized spend across hundreds of customers like WeWork, SoulCycle, Lume, and [solidcore]. Order.co has raised $75M in funding from industry-leading investors like MIT, Stage 2 Capital, Rally Ventures, 645 Ventures, and more. Order.co has been proudly named a 50 to Watch by Spend Matters and a Best Place to Work by BuiltIn and Inc. Magazine.

We are hiring a Principal Applied AI Architect to set how data and intelligence work at Order.co: the target-state architecture, the standards that keep it coherent, and the company-level bets that turn proprietary procurement data into durable advantage. This is a senior individual contributor role. You will identify company-level objectives, make the business case to fund them, and land production systems that move primary company KPIs.

You will be the Head of Data's closest technical partner: the complementary technical leader to a strategy, product, and business lens. You keep those bets grounded in what we can actually do with the systems we have, and you are the person who knows what it would take to change that.

You sit on the data team. Applied AI Scientists take the architecture into product initiatives and own the science through to the metric. Senior AI / Data Engineers take the same architecture into the platform and own the production path. You are the person who holds those two paths on one spine, and who can still get on the keyboard when the first systems need to exist.

The work is architecture-first, with a real engineering bar. You will be fluent across stacks, treat security, reliability, and maintainability as delivery criteria rather than afterthoughts, and set a production quality standard that other teams can reuse. You will also stay a working applied scientist: statistical depth, model and agent judgment, and the ability to go from a strategic imperative to a shipped system. This is not an academic role. Understanding production systems and the business they serve is required, and you will get a change through operations and the other stakeholders who gate a real deploy.

Great product capabilities stay modular when they share one canonical picture of the business: catalog, mapping, pricing, availability, and vendor truth. That picture has to be accurate first. Then it has to move: scalable pipelines and serving paths so the same data can feed the business and product independently of where they may sit. Intelligence, both deterministic and learned, is what we build on top of that advantage. The quality of the intelligence is bounded by the quality and availability of the data underneath it.

Your job is to make that stack a company strategy, not a collection of projects. Accuracy funds the platform. The platform funds intelligence. Intelligence that cannot name the data it depends on is not a bet we take.

You will define the target-state architecture for the data estate and for applied AI, aligned to business priorities rather than tooling fashion. That includes how warehouse and lake, pipelines, streaming, and integrations fit together, and how model serving, retrieval, embeddings, prompt and model management, guardrails, and evaluation fit on top of that same governed foundation. You will establish source-of-truth models and a shared semantic picture of the business so fragmented sources converge into one reliable platform, which is the foundation AI depends on.

You will set naming, modeling, lineage, and integration standards, and you will be the review that keeps new designs on that spine. You will make security, privacy, and compliance part of the design: classification, handling of sensitive data, access control, retention, and audit through every layer. Cost, performance, and reliability are first-class constraints in the architecture you defend, not optimizations someone else discovers later.

You will own where AI creates durable advantages for the data function, including build versus buy, what we sequence, and what we deliberately skip. You will make AI demand a funded platform modernization story: the data, contracts, features, and retrieval infrastructure the work requires, sized in impact, cost, risk, and return, so leadership can actually fund it. You will set institutional standards for evaluation, rollout, monitoring, model risk, and responsible AI so that no output, human or AI, ships on trust alone.

You will advise the Head of Data, and product and engineering leadership, on what the current stack can support, what it cannot, what it costs, and what it is likely to return, before the bet is made. You will set technical direction for the AI portfolio across initiatives. You will track the frontier without chasing it: pilot what matters, kill what does not, and convert the rest into production at company scale.

You will turn strategy into actionable projects, contribute to quarterly planning, and keep stakeholders in the loop as you remove blockers. You will direct a mix of human and agentic workstreams, accountable for the output of both. You will still execute: roughly half the time is implementation, because architecture that you cannot land is a slide. That includes passing the same production gates as everyone else. Operations, security, and the people who have to live with what you ship are not a handoff at the end. They are how the work gets into production.

You will mentor Applied AI Scientists and other individual contributors, and you will set the coherence standard (domain model, patterns, decision records) so someone, or an agent, can pick up work cold and land it right. You will raise the production engineering bar in AI systems across teams: observability, resilience, and the operational loop that makes a launch survivable.

• Accuracy at Scale. Make location, price, availability, and lead time trustworthy enough that buyers and product can act on them. The architecture, contracts, and quality bar that keep a single source of truth as we grow.
• Adaptive Procurement. Use that same foundation to change how customers plan and place orders, including predictive timing inside real procurement constraints: org structure, budgets, approved catalogs, and vendor contracts.
• The intelligence layer on that spine, including predictive ordering and agentic copilots in core workflows, with evaluation and operations underneath so we can tell a model that looks good offline from a capability that moves the business.

• At least 14 years in applied data science, machine learning, data architecture, or production AI systems, with a track record of landing work that moved a company-level metric you can name.
• Experience as the technical counterpart to a product, strategy, or business leader, where your job was to ground the roadmap in what the systems could actually do.
• Deep working knowledge of production systems and of the business they serve. That is required, not a plus. Research depth without a deploy record is not the profile.
• Ownership of target-state architecture across data and applied AI, not models or pipelines in isolation: you have set the spine, the standards, and the review, and you have partnered with engineering to land it.
• Repeated delivery of production systems you personally designed, defended, and got through operational and stakeholder gates to production, including serving, retrieval, evaluation, guardrails, and the operational loop around them.
• The ability to size an initiative (direct and indirect impact, cost, risk, and return) and turn that into a business case leadership actually funded.
• Real depth in current large language model and agent technology, meaning you know where it works, how it fails, how you evaluate it, and when a simpler deterministic approach is the better answer, plus a strong quantitative foundation in experimentation, statistical reasoning, and causal thinking.
• Fluency across stacks and a production engineering bar: infrastructure as code, CI/CD, observability, and security and reliability treated as delivery criteria. You can work with warehouse, lake, and cloud data platforms as production software even when you are not the person on call for every pipeline.
• Heavy daily use of AI-native engineering workflows across design, coding, debugging, and review for at least the past 18 months, along with the judgment to know where to verify and what not to trust.
• Working implementation proficiency across at least two cloud or technical ecosystems, for example AWS and GCP.
• Evidence that you have mentored senior individual contributors and set standards other teams reused, and that you can align executives and technical leaders when the situation is ambiguous.

• Experience with retrieval systems, vector search, ranking, recommendation, or production personalization.
• Experience designing source-of-truth models, semantic layers, or master and reference data that multiple product surfaces actually consumed.
• Experience setting model governance, monitoring, and responsible AI standards for an organization rather than a single initiative.
• Experience in e-commerce, B2B procurement, vendor management, financial products, or heavy integration with external systems.
• An advanced degree in a quantitative field is welcome; it is not a requirement. We care that you have shipped.

• The Head of Data is planning against a sequenced view of the data and AI portfolio that you own together, with at least one company-level bet in motion and a sized impact story behind it.
• Target-state architecture and standards are in use: new data and AI designs go through a coherent review instead of accumulating as one-off systems.
• Applied AI Scientists and Senior AI / Data Engineers are working from the same spine, so intelligence lands on governed infrastructure rather than beside it.
• Evaluation, monitoring, and responsible-AI controls are the default on production AI, not a special project.
• You have shipped something yourself, through the operational and stakeholder gates that make a deploy real. Architecture without a production artifact in this window is not the job.

You sit on the data team as the Head of Data's closest technical partner, and you are hands-on from day one. You will partner with Product, Engineering, Operations, our Applied AI Scientists, and our Senior AI / Data Engineers. We ship iteratively and we judge work by the company outcome it produces. This role is remote within the United States.

The process begins with a conversation with our recruiter, followed by a conversation with our Head of Data, who is the hiring manager for this role. If we both want to continue, we will send you a short written exercise on a real problem from our stack, designed to take about 90 minutes, and that exercise becomes the basis for a 60-minute working session with two of our senior engineers from platform and data engineering. From there you will speak with a senior business executive, our VP of Engineering, and our CTO, and we close with a values conversation with our People Ops team. That is seven conversations plus the written exercise, and your recruiter will keep you informed at each step.

• Competitive compensation package
• Employer-sponsored 401(k) with match
• Comprehensive medical, dental, and vision coverage
• Flexible time off and hybrid work environment
• The anticipated annual salary range for this role is $225,000 - $275,000. Actual compensation and title will be commensurate with experience, qualifications, knowledge, and skills.

Originally posted on Himalayas