Senior AI Platform Engineer - Customer Experience

The position

This is an architect-and-build role. Not a strategy role. Not an advisory role. You will design, engineer, and ship AI-powered systems that directly change how OPSWAT’s customers get support, solve problems, and realize value from our platform.

Our Customer Experience organization supports some of the world’s most security-conscious enterprises - organizations operating in air-gapped, OT, and mission-critical environments where a slow resolution or a missed issue has real consequences. Your job is direct and non-negotiable: use AI to reduce CX headcount through automation and digital transformation while simultaneously improving customer satisfaction. This is not a “future state” ambition - it is the explicit mission of this role.

This role operates in a matrix management structure. You will work directly with the SVP of Customer Experience and participate in CX Leadership meetings, ensuring full alignment with CX strategy and priorities. At the same time, you will report to the head of Enterprise Applications and Digital Transformation team, to ensure that your work leverage the governed platform/framework that has been developed, avoiding duplicated work and security risks. The work you do is exclusively CX-focused. Your performance will be judged on: 50% of delivery output quality from SVP of CX, and 50% on technical excellence and AI adoption by head of Digital Transformation team.

If you want to build AI systems that have a direct, measurable impact on how a world-class CX organization operates - fewer people doing manual work, more customers solving their own problems, higher satisfaction scores across the board - this role is for you.

What you’ll own

• Agentic AI Support Agent: Design and ship an agentic AI support agent that attempts to resolve incoming cases on its own - reasoning over product documentation, knowledge base articles, historical cases, release notes, and engineering knowledge to answer the customer directly. Where it can't fully resolve, it triages and prioritizes the case, gathers the right diagnostic context, and escalates to a human engineer with that context attached. It operates inside the live support workflow, takes real actions through our integrations, and learns from the outcomes of every case it touches. The goal is to reduce manual workload and shrink time-to-resolution.
• Documentation & knowledge engineering: Build AI tools that generate and continuously improve customer-facing documentation, release summaries, and self-help content. You’ll leverage LLMs, RAG architectures, vector databases, and semantic search - and close the loop by learning from real customer interactions and support trends.
• Self-service experiences for OT environments: OPSWAT’s customers aren’t typical SaaS users. Many operate in constrained, industrial, or air-gapped environments where self-service has to be purpose-built. You’ll design conversational AI interfaces and guided remediation tools tailored to those realities - not adapted from generic consumer AI patterns.
• Adoption intelligence & data tools: Build the analytics and AI-driven recommendation systems that tell us - and our customers - where adoption is strong, where it’s stalling, and what to do about it. You’ll turn platform telemetry and support data into proactive actions that improve retention and customer health.
• AI-Built Applications, Application Extensions & Business Workflows: Design and build net-new AI-powered applications and extensions to existing systems that directly improve how the CX organization operates. This includes tools such as:
  • Applications to manage and track Professional Services engagements end-to-end
  • Customer Success lifecycle tooling that surfaces health signals, renewal risk, and engagement gaps
  • Improved UI and automation layers on top of the existing Case Management application
  • Automated outreach to customers and CX personnel triggered by internal signals - such as low adoption, approaching renewals, unresolved escalations, or satisfaction dips
  The goal is not to integrate off-the-shelf tools. The goal is to build purpose-fit applications where existing platforms fall short - and extend existing ones where AI can meaningfully improve them.

What you’ll do in the first 12 months

There is no extended ramp. We expect someone who can orient quickly, identify the highest-impact opportunity, and start delivering against it within their first weeks. The team will give you the context you need - the rest is on you to move fast.

Months 1–3: Identify, build, and ship something that matters

You’ll spend the first few weeks getting deep on how our CX operation actually works - the tools, the data, the friction points. But this is not a listening tour. By the end of month three, you will have shipped something into production. It doesn’t need to be the biggest thing on the roadmap - it needs to demonstrate that you can identify a real problem, build a working AI solution, and put it in front of users. Early results are the expectation, not the exception.

Months 3–6: Prove the model, expand the foundation

With your first win in production, shift focus to measuring its impact and building on it. Harden the foundation - data pipelines, model integrations, evaluation frameworks - so that future work moves faster. Take on a second initiative with broader scope. By month six, the CX team should already be operating differently because of what you’ve built.

Months 6–12: Scale what works

Expand the platform across more use cases and customer segments. Drive measurable reduction in manual support workload, faster case resolution, and increased self-service adoption. Contribute to a playbook the team can build on well beyond your first year - and start raising the bar on what’s possible.

This role is a great fit if you...

• Have 3–8 years of software or AI/ML engineering experience and have shipped production AI systems - not just prototypes.
• Are fluent in Python and have hands-on experience with LLMs, RAG, prompt engineering, vector databases, and AI orchestration frameworks.
• Have built AI-enabled applications - chatbots, recommendation systems, automation tools - and can speak to what worked, what didn’t, and why.
• Think in outcomes, not features. You care that resolution time went down, not just that the model ran.
• Are comfortable working cross-functionally with Support, Customer Success, Product, and Engineering to identify problems before you engineer solutions.
• Bring a strong analytical foundation - you can define what success looks like, instrument it, and course-correct when the data tells you to.
• Have a solid foundation in software system design and data engineering - you understand how to architect scalable, production-grade systems, not just train models or build demos.
• Communicate with clarity