Senior Software Engineer, SpotOn

• You are the person who walks into a room, hears someone describe a painful manual process, and walks out knowing exactly how to automate it. You will:
• Own business automation end-to-end. You partner directly with operations, sales, and customer support teams to identify workflow bottlenecks and manual processes ripe for automation. You don't wait for a product manager to write you a spec. You discover the problem, scope the solution, build it, deploy it, and measure the impact
• Design and ship AI-driven systems that replace manual labor. These aren't chatbot wrappers. You're building agentic workflows and orchestration layers that handle complex, multi-step business processes autonomously — with the guardrails and error handling required for production use in a fintech environment. You build systems that people trust because they work reliably, not because someone mandated adoption
• Measure everything. You define the metrics that matter: hours recovered, error rates reduced, cycle times compressed, operational throughput gained. You report impact in business terms. "We automated X and saved Y hours per week" is the language you speak. If something isn't delivering value, the data tells you before anyone else
• Evangelize by shipping, not by presenting. Your most persuasive argument is a working system, not a slide deck. You turn skeptics into believers by showing them their own time savings. You build systems that make the whole ops org better, not just your team
• This is a high bar intentionally. We need someone who has already done this work, not someone who wants to learn
• 7+ years of professional software engineering experience — you understand how engineering and business systems actually work at scale, not just how AI demos work
• Strong proficiency in Python and TypeScript/JavaScript — you can build production systems in both. Not scripts. Not notebooks. Production
• Full-stack capability spanning backend services, APIs, and frontend interfaces. You can build the automation pipeline and the dashboard that shows it's working
• 1–2+ years working deeply with LLMs and AI agents in production — not side projects, not hackathons. Real systems handling real workflows. You can speak to failure modes, cost optimization, prompt engineering patterns, and model selection trade-offs from experience
• Built multi-agent or agentic workflows — task decomposition, agent specialization, orchestration, error recovery. You've worked with Claude Code, LangGraph, CrewAI, AutoGen, or custom agent frameworks at a level where you can articulate why you chose one approach over another
• Hands-on with AI coding tools daily — Claude Code, Cursor, GitHub Copilot, Windsurf, or similar. You don't just use them, you push their limits. You have a personal workflow that makes you measurably faster and you can teach it to others
• Comfort working directly with non-technical stakeholders. You can sit with a support lead, understand their workflow, and translate it into a technical solution without needing a translator
• You already write the majority of your code with AI assistance and you've measured the impact on your own velocity. You have opinions on which models are best for which tasks and you update those opinions as the landscape shifts
• You prototype in hours, not days. You can go from "I think we can automate this" to a working POC before end of day. You refine ruthlessly after
• You see a manual business process and your brain immediately starts decomposing it into automatable steps. You can't help it
• You communicate technical progress in business terms. You know that "we eliminated 40 hours per week of manual triage" lands better with leadership than "we built a multi-agent orchestration pipeline." You build dashboards, not just systems
• You have a healthy skepticism forged by real experience. You've seen AI fail in production. You've dealt with hallucinated outputs, flaky agent workflows, and systems that looked great in demo but broke under real load. That experience makes your guardrails better, not your enthusiasm less

Desirable

• Our operations, sales, and customer support teams run on workflows that were designed for humans — and many of them should no longer be. Manual handoffs, repetitive processes, slow queues, and duct-taped integrations are costing us speed, accuracy, and money. We're hiring a Senior AI Engineer to fix that
• Your job is simple to describe and hard to do: find the inefficiencies across our business operations and replace them with AI-driven systems that are faster, smarter, and scalable. You'll work directly with ops, sales, and support stakeholders, understand their pain points firsthand, and build the automation that eliminates them
• One week you might be streamlining a customer support triage workflow; the next, you're building an agentic pipeline that handles a multi-step process currently requiring a team of people and days of turnaround
• This is a high-autonomy, high-impact IC role. You won't have someone writing detailed specs for you. You'll get a business problem, dig into it, architect a solution, and ship it. We need someone who's been deep in AI-assisted development for years — not just prompting models, but building real orchestration layers, integrating agentic workflows into production, and producing code that's disciplined and reviewable. We are not interested in AI-generated slop
• Languages: Python, TypeScript
• Infrastructure: AWS, GitHub, CI/CD (GitHub Actions)
• Business Systems: Salesforce, internal ops tooling
• AI Stack: You'll have significant influence on tool selection — bring your opinions
• Team tooling: Jira, Slack, standard SpotOn eng stack

• Tell us about a time you looked at a manual business process, saw the waste, and built something that made it disappear. We want to hear what you automated, how you built it, and what the impact was. Bonus points if AI agents were a core part of the solution — and you can tell us where they worked, where they broke, and what you learned