My Research Workflow: Tools and Practices for PhD Students

Introduction

Over my years as a PhD student, I’ve refined a research workflow that helps me stay organized, productive, and sane. Here’s what works for me!

Paper Management

Zotero for References

I use Zotero to manage all papers I read:

• Browser extension: One-click saving of papers from arXiv, IEEE, ACM, etc.
• PDF management: Automatic download and organization
• Notes and annotations: Built-in PDF reader with highlighting
• BibTeX export: Seamless integration with LaTeX

Pro tip: Use collections to organize papers by project/topic:

📁 PhD Research
  📁 ORAM Papers
  📁 MPC Papers
  📁 Privacy-Preserving ML
📁 Background Reading
📁 Teaching Resources

Reading Strategy

Not all papers deserve equal attention:

1. First pass (5 min): Title, abstract, introduction, conclusion

   • Decide: relevant or not?

2. Second pass (30 min): Read sections, skip proofs

   • Understand main contributions and techniques

3. Third pass (2+ hours): Deep dive, understand proofs

   • Only for highly relevant papers
   • Take detailed notes in Zotero

Note-Taking System

Obsidian for Research Notes

I maintain an Obsidian vault for research notes:

research-notes/
├── Papers/
│   ├── Path-ORAM.md
│   ├── Circuit-ORAM.md
│   └── ...
├── Concepts/
│   ├── ORAM-Constructions.md
│   ├── Secure-Computation.md
│   └── ...
├── Ideas/
│   ├── Optimization-Idea-1.md
│   └── ...
└── Weekly-Notes/
    ├── 2024-W10.md
    └── ...

Key practices:

• Atomic notes: One concept per note
• Bidirectional links: Connect related ideas
• Daily research log: Track what I work on each day
• Tag system: #idea, #todo, #question, #insight

Example Note Template

# [Paper Title]

**Authors**: [Names]
**Venue**: [Conference/Journal]
**Year**: [Year]
**Link**: [URL]

## Summary
[2-3 sentence summary]

## Key Contributions
- Contribution 1
- Contribution 2

## Technical Approach
[Brief description of techniques]

## Strengths
- ...

## Limitations
- ...

## Relevance to My Research
[How this relates to my work]

## Questions
- [ ] Question 1
- [ ] Question 2

## Related Papers
- [[Link to related note 1]]
- [[Link to related note 2]]

Writing and LaTeX

Overleaf for Collaboration

For papers, I use Overleaf:

• Real-time collaboration with co-authors
• Version control via Git integration
• No LaTeX setup headaches

My LaTeX Setup

Custom macros for common notation:

% Custom commands
\newcommand{\secparam}{\lambda}
\newcommand{\negl}{\mathsf{negl}}
\newcommand{\poly}{\mathsf{poly}}
\newcommand{\Enc}{\mathsf{Enc}}
\newcommand{\Dec}{\mathsf{Dec}}

% Common notation
\newcommand{\Z}{\mathbb{Z}}
\newcommand{\F}{\mathbb{F}}
\newcommand{\G}{\mathbb{G}}

Keeps papers consistent and typing efficient!

Version Control

Even for solo writing:

git init
git add paper.tex figures/
git commit -m "Initial draft"

# Branch for major revisions
git checkout -b reviewer-response

Lifesaver when you need to recover deleted sections!

Code and Experiments

Project Organization

project-name/
├── README.md
├── requirements.txt
├── src/
│   ├── __init__.py
│   ├── oram.py
│   └── utils.py
├── tests/
│   └── test_oram.py
├── experiments/
│   ├── benchmark.py
│   └── results/
├── docs/
└── paper/

Reproducibility Practices

1. Dependencies: Use requirements.txt or conda environment.yml
2. Random seeds: Always set for reproducible experiments
3. Configuration files: YAML/JSON for experiment parameters
4. Logging: Comprehensive logging of all experiments
5. Documentation: README with setup and run instructions

Example Experiment Script

import argparse
import json
import logging
from datetime import datetime

def run_experiment(config):
    """Run ORAM benchmark with given configuration"""
    # Setup logging
    log_file = f"logs/experiment_{datetime.now():%Y%m%d_%H%M%S}.log"
    logging.basicConfig(filename=log_file, level=logging.INFO)

    logging.info(f"Configuration: {json.dumps(config)}")

    # Run experiment
    results = benchmark_oram(
        block_size=config['block_size'],
        num_blocks=config['num_blocks'],
        num_accesses=config['num_accesses']
    )

    # Save results
    results_file = f"results/{config['experiment_name']}.json"
    with open(results_file, 'w') as f:
        json.dump(results, f, indent=2)

    return results

if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument('--config', required=True)
    args = parser.parse_args()

    with open(args.config) as f:
        config = json.load(f)

    run_experiment(config)

Time Management

Weekly Planning

Every Monday morning:

1. Review previous week’s progress
2. Set 3 main goals for the current week
3. Schedule blocks for deep work (no meetings!)

Pomodoro Technique

For focused work:

• 25 minutes deep work
• 5 minute break
• After 4 pomodoros: 30 minute break

Apps I use: Forest (mobile), Focus@Will (background music)

Staying Current

Following the Field

• arXiv alerts: Daily digest for cs.CR (cryptography)
• Google Scholar alerts: For specific topics and authors
• Twitter/X: Follow researchers in my field
• Conference calendars: Track deadlines and proceedings

Reading Group

Weekly reading group with labmates:

• Rotate paper selection
• 30-minute presentation + discussion
• Great for discovering papers outside my immediate area

Communication

Academic Email

Template for reaching out to researchers:

Subject: Question about [Specific Paper/Topic]

Dear Dr. [Name],

I am a PhD student at UC Santa Cruz working on [brief description].
I recently read your paper "[Paper Title]" and found [specific aspect]
particularly interesting.

[Specific question or comment showing you read the paper carefully]

Thank you for your time and for your excellent research.

Best regards,
Apostolos Mavrogiannakis
PhD Student, Computer Science
UC Santa Cruz

Key principles:

• Be specific (shows you did your homework)
• Be concise (respect their time)
• Be professional

Mental Health

Research is a marathon, not a sprint:

• Regular exercise: 30 minutes daily (running, gym, cycling)
• Sleep: Consistent schedule, 7-8 hours
• Hobbies: Photography, hiking (totally unrelated to research!)
• Social connection: Regular hangouts with non-PhD friends

Burnout is real. Taking breaks makes you more productive, not less.

Tools Summary

Conclusion

This workflow evolved over years of trial and error. Your ideal setup will be different!

General principles:

• Automate repetitive tasks
• Document everything (future you will thank present you)
• Build systems, not just habits
• Regularly review and refine your workflow

What tools and practices work for you? I’d love to hear! Reach out via email or Twitter.

This post reflects my personal workflow as of 2024. Check back for updates as I continue refining!