Application Guide

How to Apply for Research Scientist

at FutureSearch

🏢 About FutureSearch

FutureSearch is a cutting-edge startup building AI systems that reason about the future reliably, a niche yet critical capability for strategic decision-making. Working here means tackling fundamental challenges in AI epistemics—how AI knows and validates information about the future—in a remote-first, research-driven environment.

About This Role

As a Research Scientist, you'll advance the frontier of AI forecasting by evaluating and improving large language models (LLMs) for web research and agentic reasoning. Your work will directly shape benchmarks and datasets that define how AI systems are assessed, with high impact on both the company's products and the broader AI community.

💡 A Day in the Life

Mornings might involve reviewing recent LLM papers or running experiments on a new evaluation metric. Afternoons could be spent analyzing results, writing code for a benchmark, or collaborating with the team on designing a study to test an agentic system's forecasting accuracy.

🎯 Who FutureSearch Is Looking For

  • Has a strong publication record in ML/NLP, particularly on LLM evaluation, reasoning, or knowledge assessment.
  • Hands-on experience building and analyzing benchmarks for LLMs (e.g., BIG-Bench, HELM, or custom evaluation suites).
  • Proficient in Python and familiar with experiment tracking tools (e.g., MLflow, Weights & Biases) for reproducible research.
  • Deeply curious about AI epistemics—e.g., how LLMs handle uncertainty, calibration, and temporal reasoning.

📝 Tips for Applying to FutureSearch

1

Tailor your resume to highlight projects involving LLM evaluation, forecasting, or agentic systems—mention specific benchmarks or datasets you've created.

2

In your cover letter, explicitly connect your research to open problems in assessing LLM-generated information (e.g., factuality, uncertainty estimation).

3

Showcase any experience with temporal reasoning or time-series forecasting in AI, even if not directly LLM-related.

4

Prepare a short portfolio or GitHub repo with examples of evaluation pipelines or analysis of LLM performance.

5

Mention familiarity with FutureSearch's published work or blog posts (e.g., on forecasting tournaments or AI epistemics).

✉️ What to Emphasize in Your Cover Letter

['Your specific experience designing benchmarks or evaluation metrics for LLMs.', 'Why you are excited about the intersection of AI forecasting and epistemics.', "How your research approach aligns with FutureSearch's mission of reliable reasoning about the future.", "Concrete examples of how you've improved model evaluation or identified failure modes in LLMs."]

Generate Cover Letter →

🔍 Research Before Applying

To stand out, make sure you've researched:

  • Read FutureSearch's blog posts and any published papers on forecasting and LLM evaluation.
  • Familiarize yourself with the concept of 'AI epistemics' and current debates (e.g., calibration, confidence).
  • Study the company's open-source contributions or datasets (if any) related to forecasting.
  • Understand the landscape of LLM evaluation frameworks (e.g., BIG-Bench, HELM, AlpacaEval) and their limitations.
Visit FutureSearch's Website →

💬 Prepare for These Interview Topics

Based on this role, you may be asked about:

1 Design an experiment to evaluate an LLM's ability to forecast near-term events (e.g., election outcomes).
2 How would you create a benchmark for assessing LLM-generated information reliability?
3 Discuss a paper on AI epistemics or LLM evaluation that you find impactful and its limitations.
4 How do you handle uncertainty in model predictions? Provide examples from your work.
5 Walk through a recent project where you built an agentic system or used LLMs for web research.
Practice Interview Questions →

⚠️ Common Mistakes to Avoid

  • Submitting a generic application that doesn't mention forecasting or epistemics—shows lack of research.
  • Overemphasizing traditional ML without connecting to LLM-specific evaluation challenges.
  • Ignoring the remote-first culture: failing to discuss how you collaborate and communicate asynchronously.

📅 Application Timeline

This position is open until filled. However, we recommend applying as soon as possible as roles at mission-driven organizations tend to fill quickly.

Typical hiring timeline:

1

Application Review

1-2 weeks

2

Initial Screening

Phone call or written assessment

3

Interviews

1-2 rounds, usually virtual

Offer

Congratulations!

Ready to Apply?

Good luck with your application to FutureSearch!