AI & Automation Engineer, LSO Workplace

Position summary

LSO Media Group is seeking an AI & Automation Engineer to explore how carefully designed assistance and automation could improve everyday workplace tasks for small and midsize organizations. This full-time position reports to the Director of Product & Engineering, LSO Workplace. You will own the implementation and evaluation of selected experiments, helping the team decide which capabilities are useful enough, dependable enough, and economical enough to pursue.

Potential areas of investigation include answering questions from authorized workplace information, helping administrators configure workflows, and drafting explanations of operational trends. These are exploratory directions rather than committed product features. You will combine software engineering with evaluation design, access controls, and clear human review for sensitive actions. The role requires building and validating working systems, including responsible use of AI development tools; prompting alone is not the scope of this position.

Responsibilities

• Work with the product director and customer-facing colleagues to identify specific tasks where assistance or automation could deliver measurable value.
• Build bounded prototypes using model APIs, retrieval, tool calling, or deterministic workflow logic as appropriate to the problem.
• Develop representative evaluation cases and measure task completion, factual support, failure rates, latency, and cost before making feature recommendations.
• Ensure retrieval and tool access respect tenant boundaries and user permissions, working with backend and platform engineers on enforceable controls.
• Design review and confirmation steps for consequential actions, with clear previews, limited permissions, audit records, and recovery options where feasible.
• Test for misleading outputs, prompt injection, unauthorized information access, and unsafe tool behavior; document limitations and effective mitigations.
• Compare AI-based approaches with simpler automation and recommend the approach that best meets the task's quality and operational requirements.
• Use AI coding assistants and development agents to support implementation, while reviewing generated code, validating integrations, and maintaining meaningful regression checks.
• Document experiments, model and workflow changes, monitoring needs, and release criteria so the team can reproduce results and make informed decisions.

Required qualifications

• At least four years of professional software engineering experience, including hands-on implementation of an AI-enabled application or substantial applied prototype.
• Strong programming skills in Python, TypeScript, or an equivalent language, with experience building APIs and integrating external services.
• Practical experience with model APIs, retrieval or tool use, and systematic evaluation of generated outputs against defined tasks.
• Demonstrable use of AI coding assistants or development agents, including personal ownership of code review, test design, and validation beyond the model's own assurances.
• Understanding of access controls, sensitive data handling, application threat scenarios, and the limits of probabilistic model behavior.
• A degree in computer science, software engineering, data science, or a related field, or equivalent practical experience, with clear communication of findings and uncertainty.

Preferred qualifications

• Experience deploying evaluated AI features in business software and monitoring quality after changes to models, prompts, or source data.
• Experience with permission-aware information retrieval, workflow engines, or human approval of proposed automated actions.
• Experience estimating model operating costs and improving latency or quality through caching, model selection, or simpler system designs.

Measures of success

• Experiments address clearly defined customer tasks and produce evidence that supports a proceed, revise, or stop decision.
• Evaluation results make important failure modes, operating costs, and limits visible to the product director.
• Prototypes respect organization and user permissions, with appropriate review before sensitive actions can occur.
• Changes to models or workflows can be assessed against reproducible evaluations and meaningful regression checks.

First 90 days

During your first 90 days, you will review candidate use cases with the product director, select one or two bounded experiments, and define evaluation criteria before implementation. You will build a working prototype using permitted test data, compare it with a simpler baseline, and assess quality, cost, access controls, and foreseeable failure modes. You will present a documented recommendation on what to develop further, what to revise, and what to defer, together with the controls required for any proposed pilot.

Compensation

Proposed annual base salary of $140,000–$185,000 USD. The final offer will reflect relevant experience, qualifications and the agreed scope of the position, subject to approval. This range covers base pay only; any incentive compensation or benefits would be specified separately in an employment offer.

What to include in your application

Please include a nonconfidential example of an AI-enabled system or automation you implemented, explaining the problem, your engineering contribution, the evaluation method, and a failure you identified. Also describe how you use and validate AI-assisted development; no specific AI vendor certification is required.