Hiring Applied Scientists: The Complete Guide

# Applied Scientist

**Location:** San Francisco, CA (Hybrid) · **Employment Type:** Full-time · **Level:** Senior/Staff

## About [Company]

[Company] is solving problems where standard approaches don't work. Our recommendation system serves 50M+ users, and the methods we use directly impact business outcomes. We're looking for an Applied Scientist to develop novel methods that drive real business impact.

We're a team of 80 engineers, including 6 applied scientists and 10 ML engineers. We're Series C funded ($100M) and research is a core competitive advantage-our methods are why customers choose us.

**Why join [Company]?**

- Work on problems where research matters (not just engineering)
- Access to significant compute and data
- Direct business impact from your research
- Competitive compensation with meaningful equity
- Publish papers (when appropriate for the business)

## The Role

As an Applied Scientist, you'll develop novel methods to solve complex problems in recommendation, personalization, and related domains. You'll work at the intersection of research and engineering, creating approaches that work in production at scale.

The ideal candidate values impact over publications. You're a researcher who ships-you develop novel methods and ensure they work in the real world. You're comfortable with the messiness of production data and the constraints of real systems.

## What You'll Work On

### First 30 Days
- Understand our current systems and research opportunities
- Review recent experiments and identify quick wins
- Meet with product and engineering teams to understand constraints
- Start on your first research project

### First 90 Days
- Complete your first research project with measurable impact
- Establish experimental methodology for ongoing research
- Build relationships with ML engineering team
- Contribute to research roadmap

### First Year
- Lead multiple research projects with significant business impact
- Mentor junior scientists and ML engineers
- Publish research externally (when appropriate)
- Establish [Company] as a leader in applied research

## Responsibilities

**Research (50%)**
- Develop novel methods for challenging problem domains
- Design and run rigorous experiments to validate approaches
- Stay current with research in the field
- Identify research opportunities from business problems

**Productionization (30%)**
- Collaborate with ML engineering to productionize methods
- Iterate based on production performance
- Handle production constraints (latency, cost, reliability)
- Monitor and improve deployed models

**Communication & Leadership (20%)**
- Measure and communicate business impact
- Present research to stakeholders and leadership
- Mentor junior scientists and engineers
- Contribute to research strategy

## Required Skills

- PhD in relevant field (ML, CS, Statistics) or equivalent experience
- Deep expertise in recommendation systems, NLP, or related domain
- Strong publication record or demonstrated production ML impact
- Excellent software engineering skills (production-quality code)
- Experience bringing research to production at scale
- Strong communication with non-research stakeholders

## Nice to Have

- Industry applied research experience (FAIR, DeepMind, industry labs)
- Open source contributions in ML/AI
- Experience with large-scale recommendation systems
- Publications at top venues (NeurIPS, ICML, RecSys)

## Tech Stack

**ML:** PyTorch, JAX, Ray
**Infrastructure:** Kubernetes, AWS
**Data:** Spark, Snowflake
**Experimentation:** Internal platform, MLflow

## Compensation and Benefits

**Salary:** $200,000 - $280,000 (based on experience)
**Equity:** 0.08% - 0.20% (standard 4-year vesting)
**Bonus:** 15-20% annual target

**Benefits:**
- Health, dental, vision (100% covered for employee and dependents)
- 401(k) with 4% match
- Unlimited PTO (average taken: 4 weeks)
- $5,000/year conference and research budget
- Publication bonus program
- Access to compute for personal research projects

## Interview Process

1. **Recruiter Screen** (30 min) - Background and mutual fit
2. **Research Presentation** (60 min) - Present your past research
3. **Technical Interview** (60 min) - ML fundamentals and coding
4. **Onsite** (4 hours, can be virtual)
   - Research Deep Dive (60 min) - Discuss your research in depth
   - System Design (60 min) - Design an ML system for a real problem
   - Productionization (45 min) - How you work with engineering
   - Culture Fit (45 min) - Team fit and values
5. **Offer** - Within 48 hours

We aim to complete the process in 2-3 weeks.