About Costory:

• We’re building the FinOps Assistant we wish we had as large cloud clients. After managing multi-million dollar budgets across GCP, AWS, and Datadog, we know how painful cloud cost management can be. FinOps often falls on senior engineers, forced to manually monitor costs and investigate spikes with little context.
• We believe FinOps and engineering teams shouldn’t waste time digging through billing consoles and dashboards. LLMs alone won’t solve anomaly detection or cost allocation—the real challenge is adding context to messy billing data. That’s why we’re building an agent that correlates costs with technical and product events, explains changes, and surfaces insights directly where teams work, so they can focus on optimization.
• Long-term, our vision is to shift FinOps left, just like SecOps. Engineers should understand cost impact as they build, not after the fact. Costory becomes their co-pilot—helping them make smarter trade-offs between performance, reliability, and cost early in the development cycle.

Your Mission:

As a ML engineer / data science Intern, you'll be at the forefront of developing and enhancing our investigation engine. You'll work on projects that directly impact our product's capabilities, gaining hands-on experience with time series analysis, machine learning, and large language models (LLMs). This internship is perfect for a student in their final year or taking a gap year, looking to apply their skills in a dynamic, real-world setting.

What You'll Be Working On:

• Project 1: Advanced Time Series Anomaly Detection:
  • Challenge: Move beyond basic time series analysis. To improve cost anomaly detection through enhanced noise reduction and the integration of business/technical metrics, we will evaluate traditional time series forecasting against advanced deep learning models.
  • Your Role: Research, implement, and evaluate different time series models (deep learning based / ARIMA / Prophet). Develop a methodology to evaluate the model.
• Project 2: Intelligent Reporting:
  • Challenge: Transform complex data analysis into clear, concise, and actionable reports tailored to different stakeholders.
  • Your Role: Leverage large language models (LLMs) to automate the generation of insightful summaries from our FinOps investigation engine. Design and implement a system that produces reports of varying lengths and styles, optimized for different audiences (finance, tech, product) and communication channels (Slack, email, Teams).

Your Toolkit (Our Tech Stack):

• Google Cloud Platform (GCP)
• BigQuery (Data Warehousing)
• dbt (Data Build Tool) & Cloud Workflows (Orchestration)
• Node.js (Express) & React (Frontend) with TypeScript

What We're Looking For:

• A student in their final year or taking a gap year in Computer Science, Data Science, Mathematics, or a related field.
• Strong analytical and problem-solving skills.
• Familiarity with time series analysis, machine learning, or data analytics is a plus.
• Experience with Python, SQL, or JavaScript/TypeScript is beneficial.
• A proactive, curious, and collaborative mindset.