SpektrBot is an AI-backed chat assistant for Amazon Ads designed to help business quickly access and analyze advertising data without needing technical skills or extensive support. It simplifies the process of retrieving business insights by allowing users to ask questions in natural language, enabling them to gain answers about ad performance, revenue, and other key metrics efficiently. By streamlining data exploration, SpektrBot enhances decision-making for various teams, such as Finance, Product Management, and Sales, reducing the time spent searching through reports and dashboards.

As a Software Engineer at SpektrBot, you will play a pivotal role in shaping the future of our AI-backed chat assistant designed to streamline access to Ads data and insights for business users. You will drive innovation and facilitate the adoption of cutting-edge Generative AI technologies within the advertising space.

Key job responsibilities
- Drive the design, development, and continuous improvement of machine learning models powering SpektrBot, with a focus on enhancing answer accuracy, contextual relevance, and user trust.
- Lead experimentation with Generative AI techniques—including prompt engineering, fine-tuning, and metadata refinement—to improve response quality and reduce hallucinations.
- Architect and implement scalable, low-latency inference pipelines to support real-time, production-grade GenAI applications.
- Own the end-to-end execution of ML solutions, ensuring they meet high standards of quality, performance, and reliability while delivering meaningful business insights.
- Collaborate cross-functionally with Scientists, Finance, Business Analysts, Product Managers, and Sales Leaders to translate complex business needs into innovative ML solutions.
- Champion rapid iteration and experimentation, contributing to the continuous enhancement of SpektrBot and other self-service analytics tools.
- Mentor junior engineers in applied ML and GenAI best practices, code quality, system design, and debugging across the full software development lifecycle.