Applied AI Developer

• This role requires 4days/week in our NYC (FiDi office) About Lorum Global payments are not broken. Incentives are. Clearing has been deprioritized inside balance sheet driven institutions whose models rely on lending and interest. When liquidity takes priority over settlement, payments slow and certainty drops. The same financial institutions that distort clearing as providers are disadvantaged as users. They are forced into fragmented setups, inconsistent rails, duplicated compliance, and unpredictable timelines. Stablecoin shortcuts and treasury pooling treat symptoms at the surface, but almost no one is rebuilding the underlying infrastructure in each market. Rebuilding clearing from the ground up We are rebuilding clearing as its own specialist function. We act as a clearing and transaction banking partner for regulated institutions, with treasury built into the core so liquidity, settlement, and reconciliation sit in one controlled system. Our platform unifies global and local licenses, direct central bank clearing, and access to domestic rails. We allow clients to open named customer accounts in every market we operate, collecting funds and paying out through a single network while retaining full ownership of their customer relationships. Market expansion becomes as simple as one correspondent relationship, not hundreds. Why Lorum Joining Lorum means contributing to one of the most ambitious clearing infrastructure projects in global finance. You will help shape settlement systems that perform under real regulatory standards and institutional volumes. You will build for regulated institutions that rely on precision, predictable timelines, and regulatory integrity. It is about working across currencies, markets, and supervisory frameworks to deliver reliable, final settlement. Role purpose We are looking for a pragmatic, hands-on Applied AI Developer to join our growing AI team within the office of the CEO. Reporting to the Applied AI Solutions Lead, you will work directly against a live backlog of real business problems — sourced from across Sales, Operations, Compliance, and Platform — and ship AI-driven tooling that gives time back to the people doing the work. This is a builder's role. Your output will be measured in working software in production, not strategy decks. You will prototype quickly, validate impact with the teams who use what you build, and iterate. The best fit views AI as a powerful addition to the software engineering toolkit — useful for automating the mundane so the rest of the company can focus on the meaningful. Key Responsibilities • Backlog execution: Pull problems from the Applied AI intake and turn them into shipped tools — internal apps, agentic workflows, RAG-powered Q&A, automated reporting, and integrations into the systems Lorum already runs on. • Rapid prototyping: Get a working MVP in the hands of users within days, not quarters. Validate business impact before scaling. • System integration: Build the connective tissue between our internal data sources (Slack, Notion, Drive, GitHub, our core platform) and AI models — securely, observably, and in line with our risk posture. • Evals and monitoring: Implement evaluations so we know AI outputs are accurate, safe, and actually saving time rather than creating new work. • Embedded delivery: Pair directly with the teams you are building for — Sales, Ops, Finance, Compliance, Engineering — so the tools fit how people actually work. Ideal candidate Must-haves • Software engineering: Strong proficiency in Python and/or TypeScript, with experience shipping production web apps (FastAPI, Next.js, or similar). • AI in production: Hands-on experience building with LLM APIs (Claude, OpenAI, Gemini, or similar), including tool use, structured outputs, and retrieval-augmented generation. • Pragmatic stack sense: Comfort working across vector stores (pgvector, Pinecone, Weaviate, or similar) and modern deploy targets (Vercel, Railway, Supabase, or similar). • Ship-it mindset: A track record of getting working software in front of users quickly. Bias to iteration over architecture astronautics. • Human-in-the-loop instinct: Ability to design tools and UIs where AI assists humans transparently, rather than operating as a black box. Nice-to-haves • Experience in financial services, fintech, or another regulated domain. • Familiarity with agentic frameworks and patterns, whether direct orchestration or libraries (LangGraph, CrewAI, and the like). • Background in another engineering discipline, data science, or quant — anywhere shipping mattered more than the title. Location and compensation • New York City based in our FiDi office, 4 days/week with flexibility • USD 160,000 – 190,000 base, plus employee stock ownership (ESOP) Benefits • Opportunity to travel (if applicable) • Flexible vacation policy • Private healthcare • Employee stock ownership (ESOP) • Flexible working and autonomy • Pay it