// Current projects:
- Vault Keeper (AI accountability system)
- Building a neural network from scratch
- Automated trading bot (work in progress)
- Various web scraping utilities
- Machine learning experiments
- This portfolio site
- Calorie tracking automation
- README side project

// Tech stack:
Python, JavaScript, React, Node.js, PostgreSQL, Docker
                

NEURAL NETWORKS FROM SCRATCH
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Building a neural network library in Python without using
existing ML frameworks like TensorFlow or PyTorch.

FEATURES
- Custom backpropagation implementation
- Multiple activation functions (ReLU, Sigmoid, Tanh)
- Various optimization algorithms (SGD, Adam, RMSprop)
- Layer types: Dense, Convolutional, Pooling
- Loss functions: MSE, Cross-entropy
- Training visualization and metrics

GOALS
- Deep understanding of neural network internals
- Learn gradient descent and backpropagation math
- Implement from NumPy arrays up
- Compare performance with established frameworks

CURRENT STATUS
In progress - implementing convolutional layers

Technologies: Python, NumPy, Matplotlib
                

AUTOMATED CRYPTO TRADING BOT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Algorithmic trading bot for cryptocurrency markets using
technical analysis and machine learning predictions.

FEATURES
- Real-time market data streaming
- Technical indicators (RSI, MACD, Bollinger Bands)
- ML model for price prediction
- Risk management and position sizing
- Backtesting framework
- Paper trading mode for testing strategies
- Multi-exchange support

STRATEGIES
- Trend following with momentum indicators
- Mean reversion on volatility
- ML-based pattern recognition
- Sentiment analysis from social media

CHALLENGES
- Market volatility and risk management
- Latency and execution speed
- Overfitting in backtests
- API rate limits

Technologies: Python, Pandas, scikit-learn, WebSocket APIs
                

WEB SCRAPING TOOLKIT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Collection of web scraping utilities for data extraction,
monitoring, and automation across various websites.

CAPABILITIES
- Dynamic content scraping (JavaScript-rendered pages)
- Anti-bot detection evasion
- Rotating proxies and user agents
- CAPTCHA solving integration
- Rate limiting and politeness
- Data cleaning and normalization
- Export to CSV, JSON, databases

USE CASES
- Price monitoring and comparison
- Research data collection
- Job listing aggregation
- Social media analytics
- News and content aggregation

TECHNICAL APPROACH
- Selenium for dynamic content
- BeautifulSoup for HTML parsing
- Scrapy framework for large-scale scraping
- Headless browser automation
- Async requests for performance

Technologies: Python, Selenium, BeautifulSoup, Scrapy, PostgreSQL
                

INTERACTIVE PORTFOLIO SITE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

This portfolio website featuring a 3D brain-shaped knowledge
graph visualization of skills and technologies.

FEATURES
- 3D rotating brain visualization
- Canvas-based particle physics
- 48+ nodes representing skills/technologies
- Interactive node exploration
- Drag to rotate, scroll to zoom
- Light/dark mode toggle
- Minimizable graph panel
- Responsive design

TECHNICAL HIGHLIGHTS
- Custom 3D projection mathematics
- Brain-region based node positioning
- Floating animation with sine waves
- High-DPI canvas rendering
- CSS custom properties for theming
- No frameworks - vanilla JavaScript

BRAIN REGIONS
- Frontal: Technical skills & programming
- Parietal: Tools & infrastructure
- Temporal: Neuroscience & learning
- Occipital: Frontend & communication
- Central: Core identity
- Cerebellum: Experimental projects

Technologies: Vanilla JavaScript, HTML5 Canvas, CSS
                

VAULT KEEPER - AI ACCOUNTABILITY SYSTEM
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

An Obsidian-based personal accountability system powered by
Claude AI that reads daily journals and holds you accountable
to your own stated goals.

CORE CONCEPT
Uses your own words against you. Reads journal entries, tracks
behavioral patterns, and generates accountability briefs that
call out the gap between what you say you'll do and what you
actually do.

FEATURES
✓ Pattern Detection
  - Self-sabotage cycles
  - Avoidance behaviors (gym, scrolling, nutrition)
  - Late Vyvanse timing patterns
  - Post-success drift windows

✓ Data Analysis
  - Correlation finder (sleep→exam performance, etc.)
  - Weekly trend analysis
  - Momentum tracking with streak gamification
  - Predictive drift warnings

✓ Accountability Tools
  - Daily morning check-ins
  - Friday comprehensive reviews
  - One-button reset plans when drift activates
  - Priority Stack dashboard with live metrics

✓ Smart Features
  - Pattern prediction (which pattern activates next)
  - Visual tracking (charts, graphs, heatmaps)
  - Tag management for vault organization
  - System innovation suggestions

TECHNICAL ARCHITECTURE
- Obsidian vault as data source
- Claude Code CLI integration
- Custom slash commands (/morning, /friday-review, /reset)
- Specialized AI agents (vault-keeper, pattern-scan, etc.)
- MCP skills system for modular functionality
- Markdown-based outputs for vault integration

WHY IT WORKS
Traditional accountability = external pressure
Vault Keeper = your own documented reality

When you write "I'll wake up at 7am" then sleep til 10am for
3 days straight, VK shows you the pattern. No judgment, just
facts from your own journal.

STATUS
✅ Complete and ready to distribute
Production-ready accountability system for Obsidian users

USE CASES
- Students tracking study habits and exam performance
- Anyone with ADHD managing medication timing and routines
- Fitness tracking and pattern recognition
- General productivity and goal accountability

Technologies: Claude AI, Obsidian, Markdown, MCP Protocol
Contact: jake@bryt.works | jakedev.org
                

AUTOMATED CALORIE TRACKING
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Integration system to automatically populate nutrition
metrics in Obsidian vault from external calorie tracking
sources.

CONCEPT
Eliminate manual data entry for nutrition tracking by
pulling calorie/macro data directly into daily journal
templates.

PLANNED FEATURES
- API integration with MyFitnessPal, LoseIt, or Cronometer
- Automatic daily nutrition summary in journal
- Macro breakdown visualization
- Correlation with energy levels and productivity
- Integration with Vault Keeper for pattern analysis

TECHNICAL APPROACH
- REST API calls to nutrition tracking services
- Data parsing and formatting
- Obsidian Dataview integration
- Scheduled data pulls (daily/hourly)
- Local caching for offline access

VAULT KEEPER SYNERGY
Once integrated, VK can analyze:
- Nutrition → Vyvanse effectiveness correlation
- Protein intake → gym performance
- Calorie deficit → energy/mood patterns
- Meal timing → sleep quality

STATUS
💡 Idea phase - not yet implemented

CHALLENGES TO SOLVE
- API access and authentication
- Data format standardization
- Handling missing/incomplete data
- Privacy and local-first approach

Technologies: Python, REST APIs, Obsidian Dataview
                

README PROJECT
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Side project started post-exam period (mentioned Nov 5).

STATUS
🚧 Work in progress

DESCRIPTION
[Details not in vault - more information needed]

This appears to be a project you started working on after
finishing exams, during a period of high motivation.

TIMELINE
- Started: Around November 5, 2024
- Current status: Unknown
- Context: Post-exam side project

NOTES
To expand this section, please provide:
- What is README?
- What problem does it solve?
- What technologies are you using?
- Current progress and goals?

Technologies: [To be determined]
                

Jake Davis

Software developer and neuroscience student.
Interested in machine learning, web development, automation, and the human brain.

Currently working on various personal projects and learning new technologies.
Always looking for interesting problems to solve.

When I'm not coding, you'll find me:
- Photography
- Outdoors (hiking, backpacking, mountaineering)
- Climbing (bouldering and lead)
- Skiing
- Traveling

Location: Earth
Status: Online
                

// Get in touch:

email: jake@bryt.works
github: github.com/quakeob
linkedin:
instagram: @jjakedavis

// Or just send a message:
echo "hello" | mail jake@bryt.works
                

JAKE DAVIS
Software Developer & Process Engineer

EXPERIENCE
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Peaks Lighting
Process Engineer – Aluminum Manufacturing | 2025 – 2026
Oversees and optimizes production processes for aluminum lighting
track systems. Responsible for improving fabrication efficiency,
ensuring product quality, and implementing engineering solutions
across manufacturing and warehouse operations. Collaborates with
design, fabrication, and logistics teams to streamline workflows,
reduce waste, and enhance overall production performance.

⸻

Self-Employed
Programmer / Software Developer | 2021 – 2025
Developed custom software tools, websites, and automation systems
for small businesses and creative projects. Managed full development
cycles—from concept and design to deployment—while maintaining
client relationships and technical documentation.

Show Setup & Operations Coordinator | 2024 – 2025
Assisted in the setup, organization, and execution of large-scale
drone light shows. Coordinated logistics, equipment preparation,
and site layout for synchronized aerial performances. Collaborated
with technical teams to ensure precision, safety, and visual
consistency throughout each production.
droneshow.com

⸻

Burton – Park City
Retail Sales Representative | 2020 – 2021
Provided expert guidance on snowboarding gear, outerwear, and
accessories for a leading winter sports brand. Delivered high-level
customer service, assisted with store merchandising, and maintained
product knowledge to enhance customer experience.

⸻

Ice Castles
Photo & Video Team | 2017 – 2020
Captured and produced high-quality visual content showcasing
large-scale ice structures and visitor experiences. Coordinated with
creative and marketing teams to deliver media used in national
promotional campaigns and event branding.
icecastles.com
Featured on TODAY Show

Winter Build Crew | 2018 – 2020
Constructed large-scale ice installations and managed structural
integrity during winter operations. Worked in extreme environments
to design, maintain, and sculpt ice features, ensuring guest safety
and visual appeal.

EDUCATION
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Neuroscience | Brigham Young University | 2027 (Expected)
Relevant Coursework: Neurobiology, Cognitive Science, Statistics

SKILLS
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Languages: Python, JavaScript, Java, C++, Spanish
Frontend: React, HTML/CSS, TypeScript
Backend: Node.js, Express, PostgreSQL
Tools: Git, Docker, AWS, Linux
Manufacturing: Process optimization, quality control, workflow design
Creative: Photography, videography, drone operations
                

[2025-01-11 23:45] Added ASCII animation easter egg - Bad Apple in the terminal
[2025-01-11 22:30] Graph auto-minimizes now when viewing content. UX win.
[2025-01-11 21:15] Implemented connection highlighting. Hover reveals relationships.
[2025-01-11 20:00] Brain regions are now filterable. Frontal lobe = programming skills.
[2025-01-11 19:00] Physics controls added. Nodes spread and float speed adjustable.
[2025-01-11 18:00] Auto-rotate re-enables after 4s of inactivity. Subtle but nice.
[2025-01-11 16:30] Added Vault Keeper project. Personal accountability system.
[2025-01-11 15:00] Expanded knowledge graph to 80 nodes. Every project mapped.
[2025-01-11 14:00] Built 3D brain visualization from scratch. No frameworks.
[2025-01-11 12:00] Started portfolio rebuild. Terminal aesthetic + interactive graph.
[2024-01-15 14:32] Starting new ML project
[2024-01-12 09:15] Coffee intake: excessive
[2024-01-10 23:47] Fixed the bug. It was a semicolon.
[2024-01-08 16:20] Why do I always code better at night?
[2024-01-05 11:30] New year, new repository structure
[2024-01-03 08:45] Reading papers on transformer architectures
[2024-01-01 00:01] Hello 2024
                

Loading...