Datum

Building the internet for AI

Company

Why We Exist

We’re convinced that basically every important business going forward will be digital, and will require “hyperscaler” or “modern service provider” advantages to truly compete and win. However, most companies simply don’t have the capabilities, the resources, the relationships, or the time to do it all. We believe we can help 1-2k of them play to win. This is our “why”.

Over the last cycle, developers were abstracted from the lowest, most critical layers of networking. Powerful capabilities like private connectivity, peering, and deterministic routing are now basically invisible to them, with a high barrier to entry and an interface stuck in the telco era. At the same time, our software-driven world is getting more complex and fragmented right alongside a rapid increase in the value of connected, data-aware global experiences.

We think developers need networking superpowers to build the best version of the future and connect securely with the partners, providers, applications, and customers that help them win. Our mission is to make it happen.

What We Believe

Open is better - We believe in the power of open source to build trust, enable innovation, and serve the “non bell curve” crowd. Our products are backed by AGPLv3 licenses so users can invest confidently and adapt as needed.

Software is the customer - We design for developers who think in code, not infrastructure engineers stuck in the telco era. Everything should be programmable, accessible, and AI-native.

Partners have real value - We’re not trying to replace everything. We’re building a neutral ecosystem where the best providers can thrive and connect with their customers.

Differentiation matters - While public clouds focus on 80% “bell curve” use cases, we serve the providers who need to drive differentiation in price, performance, features, timing, and reach.

Our Values

We are connectors — of people, businesses, apps and networks. At our core, we believe in creating bridges where others build walls.

We are operators at heart who know how to get stuff done. We don’t just theorize; we build, ship, and scale.

We value grit, humility and hunger - We welcome bold ideas and reject inertia, but also embrace empathy and continuous learning.

We invest in relationships - We operate with honesty and look for savvy yet humble people who can’t help but act like owners.

Who We Are Building For

Our customers are AI-native providers and digital leaders: AI startups, alt clouds, and next-gen enterprises who are building what’s next. They’re focused on innovation, speed to market, and making the future possible.

These are companies like Vercel, Snowflake, Roblox, Palo Alto Networks, Hugging Face, Datadog, American Express, Databricks, and Walmart - providers who want to drive differentiation or meet increasingly complex regulatory requirements.

They need to:

Acquire “hyperscaler” infrastructure advantages
Efficiently operate connected cloud experiences at scale, globally and across diverse environments
Securely connect a fragmented ecosystem: 10’s of cloud regions, 100’s of partners, 1000’s of customers
Have end-to-end observability across clouds and providers, using their standard tooling
React to increasingly complex regulatory, security, privacy, performance, and business requirements
Evolve alongside AI and interact with infrastructure using reasoning and AI-native concepts

Our Open Source Strategy

We’re building Datum Cloud as open source with AGPLv3 licensing because we believe transparency and community are essential for infrastructure that people depend on.

Why open source matters for us:

Trust - Our customers can see exactly what they’re running and how it works
No lock-in - You can run Datum Cloud anywhere - our cloud, your cloud, or on-premises
Community innovation - The best solutions come from diverse contributors solving real problems
Neutrality - Open source reinforces our position as a neutral operator, not a competitor

Our open source commitment:

Core Datum Cloud functionality is fully open source
Community contributions are welcomed and celebrated
Documentation and examples are extensive and maintained
We provide clear paths for both self-hosted and managed deployments

By 2027, we aim to have 150+ open source contributors actively improving the platform.

Our Neutral Strategy

Unlike hyperscale clouds that want to lock you into their ecosystem, we’re building Datum to be genuinely neutral. We succeed when our ecosystem partners succeed.

What neutral means:

No margin stacking - We use transparent, rebill-style commercial models
Partner-first mindset - We’re designed to amplify your differentiation, not compete with it
Open ecosystem - We connect providers with their unique ecosystem advantage
Technology agnostic - We support diverse technology partners and their channels

Why neutrality is our strategy:

The market needs a “Switzerland” for digital infrastructure
Providers need a place to collaborate without competitive dynamics
Innovation happens when different technologies can work together seamlessly
Trust is earned by consistently acting in everyone’s best interest

Our AI Strategy

We’re not just building for the current internet - we’re building the internet that AI needs.

AI is changing everything:

Complexity explosion - AI workloads are more distributed, data-intensive, and performance-sensitive
New interfaces - AI agents need programmatic, reasoning-friendly interfaces to infrastructure
Ecosystem importance - AI leaders win by orchestrating powerful ecosystems, not building everything themselves
Global requirements - AI services need to run everywhere with deterministic performance

How we’re building for AI:

AI-native interfaces - Our platform speaks the language of modern AI systems and agents
Intelligent routing - AI-driven traffic optimization and policy enforcement
Ecosystem orchestration - Seamless connection between AI providers, cloud platforms, and enterprise customers
Performance guarantees - Deterministic routing and observability for mission-critical AI workloads

Our 10-year target: The most trusted way for AI and digital ecosystems to interact (with millions of active connections to prove it).

How We Got Here

Datum was born from the experience of building and scaling infrastructure companies like Packet (acquired by Equinix). We saw firsthand how the best teams get stuck stitching together complex infrastructure instead of focusing on their core differentiation.

The founding insight: What if we could give every innovative company the networking superpowers that only hyperscalers have today?

Key moments in our journey:

Recognized the growing abstraction gap between developers and networking
Understood that AI would accelerate this complexity exponentially
Saw the opportunity to update the collaborative, open architecture of the original internet
Decided to build neutral infrastructure that serves providers, not competes with them
Committed to open source as the foundation for trust and community

What Inspires Us

The original internet - We’re inspired by the collaborative, open architecture that built the foundation of our digital world. We want to update that vision for the AI era.

The next 1,000 hyperscalers - Every important business needs infrastructure superpowers to compete. We want to democratize access to those capabilities.

Developer empowerment - The best solutions come from empowering smart people to solve their own problems, not prescribing one-size-fits-all solutions.

Ecosystem thinking - The companies that win long-term sit at the center of powerful, value-creating ecosystems.

How We Make Money

We make money when our ecosystem thrives. Our business model aligns our success with our customers’ success.

Revenue streams:

Datum Cloud - Managed infrastructure services with transparent, usage-based pricing
Datum OS - Business operating system for product-led companies
Datum Marketplace - Neutral commercial ecosystem with rebill-style margins

Pricing philosophy:

Transparent, predictable pricing that scales with value delivered
No margin stacking on partner solutions
Free tiers that let developers experiment and learn
Enterprise pricing that reflects the business value we create

2025 targets:

$25k MRR
500 quality organizations on the platform
5 AI native companies “leaning in”

2027 vision:

$2M MRR
70% gross margin
120% net dollar retention

What Are Our Rituals

Weekly rhythms:

Team syncs - Each small team meets weekly to review progress and plan ahead
Demo Friday - We show what we’ve built and celebrate progress
Community office hours - Regular time for open source contributors and users

Monthly rhythms:

All-hands - Company-wide updates and strategic discussions
Customer feedback reviews - Deep dives into user feedback and feature requests
Ecosystem partner check-ins - Regular touchbases with key partners

Quarterly rhythms:

Planning cycles - Set goals and priorities for the next quarter
Board meetings - Review metrics, challenges, and strategic decisions
Team offsites - In-person time for relationship building and strategic work

How We Talk to Each Other

We default to transparency - Most communication happens in public channels where anyone can contribute or learn.

We write things down - Important decisions, context, and knowledge live in our handbook, RFCs, and documentation, not just in someone’s head.

We give direct, kind feedback - We care enough about each other and the company to have hard conversations when needed.

We assume positive intent - We start from the assumption that everyone is trying to do their best work and help the company succeed.

We communicate for our global team - With team members across time zones, we prioritize clear, written communication that works asynchronously.

Deciding What Products to Build

Our product decisions flow from our core focus: connecting AI-native providers to their unique ecosystem advantage.

Decision framework:

Does it help providers differentiate? - We build features that unlock competitive advantages
Does it strengthen the ecosystem? - We prioritize capabilities that create network effects
Can we be the best at it? - We focus on areas where we have unique expertise and positioning
Is it AI-native? - We design for the interfaces and workflows that AI systems need

Current product priorities:

Datum Cloud - Open network cloud with global reach and AI-native interfaces
Datum OS - Business operating system for product-led companies
Datum Marketplace - Neutral commercial ecosystem for frictionless partnerships

2025 product goals:

Solid v1 of Datum OS available to external users
Five public/specialty cloud provider integrations
Three magnetic AI providers onboarded
One incumbent partner (security, data, networking) in place

Where We Are Now

Current state (2025):

Early-stage company with proven founding team
MVP of Datum Cloud and Datum OS in development
Growing community of early adopters and design partners
Open source foundation being built in public
Strategic partnerships being established

Key metrics we’re tracking:

Monthly recurring revenue growth
Number of quality organizations on platform
Open source contributor growth
Ecosystem connections created
Customer satisfaction and retention

How You Can Help

If you’re a developer or infrastructure operator:

Try our open source tools and give us feedback
Contribute to our repositories
Share your infrastructure challenges with us
Join our community discussions

If you’re building an AI-native company:

Partner with us as a design partner
Connect us with others facing similar challenges
Help us understand your infrastructure needs
Consider Datum for your networking requirements

If you’re a technology provider:

Explore partnership opportunities
Join our neutral marketplace
Contribute to our open source ecosystem
Help us serve mutual customers

If you’re interested in joining the team:

Check out our open positions
Contribute to our open source projects
Engage with our community
Share your ideas for building the internet for AI

People

How We Work

We’re a remote-first company that believes in async-first communication and autonomy. Every team member is empowered to make decisions in their domain and drive projects forward.

Core principles:

Remote-first - We’re designed for distributed work across time zones
Async-first - We default to written communication and recorded decisions
Autonomy with alignment - Clear goals and context, with freedom to execute
Transparency by default - Information is open unless there’s a specific reason to keep it private

How We Think About Remote Work

Remote work isn’t just about where we sit - it’s about how we design processes, communication, and culture to work better for everyone.

Benefits of remote:

Access to global talent pool
Focus time without office distractions
Flexibility to work when and where you’re most productive
Reduced commute stress and environmental impact
Better work-life integration

How we make remote work:

Clear, written communication standards
Overlap hours for real-time collaboration when needed
Regular video calls for relationship building
Annual team gatherings for in-person connection
Tools and processes designed for distributed teams

What Are the Benefits

Health & Wellness:

Comprehensive health insurance
Mental health support
Fitness/wellness stipend

Work Setup:

Home office stipend for equipment
Co-working space allowance
High-quality laptop and peripherals

Time & Flexibility:

Unlimited PTO (with minimum requirements)
Flexible working hours
Parental leave
Conference and learning budget

Financial:

Competitive equity package
Regular salary reviews
Transparent compensation bands

How to Spend Money

We trust team members to spend money responsibly on behalf of the company.

General principles:

Act like an owner - spend company money like it’s your own
When in doubt, ask - but don’t let perfect be the enemy of good
Document expenses clearly for transparency
Choose quality tools that help us work better

Spending guidelines:

Under $500: Just do it (within reason)
$500-2000: Get manager approval
Over $2000: Needs finance team review
Recurring subscriptions: Always get approval first

What Is the Travel Policy

Business travel:

Book travel that balances cost and convenience
Choose reasonable accommodations (good quality, convenient location)
Expense meals, transport, and reasonable incidentals
Use company credit card when possible

Team gatherings:

We’ll coordinate and book accommodations for team events
Travel costs are fully covered
We choose locations that minimize total travel time/cost

Conference travel:

Pre-approval required for conferences
Budget includes registration, travel, accommodation, meals
Share learnings with the team afterward

How to Give Feedback

Feedback is a gift - both giving and receiving it well is essential for our growth.

Giving feedback:

Be direct but kind
Focus on specific behaviors and outcomes
Offer suggestions for improvement
Give feedback promptly when relevant
Use “I” statements to avoid blame

Receiving feedback:

Listen without getting defensive
Ask clarifying questions
Thank the person for their input
Follow up on actions you commit to

Formal feedback:

Regular 1:1s with your manager
Quarterly goal reviews
Annual 360-degree feedback process
Project-specific feedback sessions

How We Recognize Peers

Recognition reinforces our values and celebrates great work.

Ways we recognize contributions:

Shout-outs in team channels for great work
Peer nominations for going above and beyond
Quarterly awards for living our values
Professional development opportunities for top performers
Increased responsibility and autonomy for proven contributors

What we celebrate:

Exceptional customer outcomes
Innovative technical solutions
Outstanding collaboration
Living our values under pressure
Mentoring and helping teammates grow

How We Think About Titles

Titles should reflect what you actually do, not just seniority. We prefer descriptive titles that help people understand your focus and expertise.

Title philosophy:

Clarity over hierarchy
Function over seniority
Authenticity over inflation
Evolution as roles grow

Examples:

“AI Infrastructure Engineer” not “Senior Engineer”
“Developer Experience Lead” not “Manager”
“Ecosystem Partnerships” not “Business Development”

Engineering

What Is Our Tech Stack

Core infrastructure:

Go - Primary backend language for performance and reliability
Rust - For performance-critical networking components
TypeScript/React - Frontend applications and dashboards
PostgreSQL - Primary database for structured data
Redis - Caching and session management

Infrastructure:

Kubernetes - Container orchestration
Terraform - Infrastructure as code
BGP/BIRD - Network routing protocols
DPDK - High-performance packet processing
eBPF - Network observability and security

AI/ML:

Python - ML workflows and data processing
PyTorch - Machine learning models
CUDA - GPU acceleration where needed

How We Use RFCs and Enhancements

We use RFCs (Request for Comments) to design significant technical changes collaboratively.

When to write an RFC:

New product features or major changes
Architecture decisions that affect multiple systems
Breaking changes to APIs or interfaces
Complex technical problems requiring research

RFC process:

Write the RFC document with problem, solution, alternatives
Share for async feedback and iteration
Schedule RFC review meeting if needed
Get approval from relevant stakeholders
Use RFC as implementation guide

How We Plan Our Cycles

We work in 6-week cycles with 2-week cool-downs, inspired by Basecamp’s Shape Up methodology.

Cycle structure:

6 weeks building - Focus on one major project per person/team
2 weeks cool-down - Bug fixes, exploration, planning next cycle
Planning week - Shape work for next cycle, assign teams

Planning principles:

Shape work before assigning it
Give teams uninterrupted focus time
Solve real customer problems
Iterate based on usage and feedback

How We Ship New Features

Development process:

Design - Create detailed specs and get feedback
Build - Implement with automated tests
Review - Code review and testing
Deploy - Gradual rollout with monitoring
Measure - Track usage and impact

Quality gates:

All code must have tests
Security review for sensitive changes
Performance testing for infrastructure changes
Documentation updated before shipping

How We Review Pull Requests

Code review is essential for code quality, knowledge sharing, and catching issues early.

Review checklist:

Does the code solve the right problem?
Is it well-tested and documented?
Does it follow our coding conventions?
Are there any security or performance concerns?
Could it be simpler or more maintainable?

Review process:

All PRs need at least one approval
Complex changes need domain expert review
Security-sensitive changes need security team review
Breaking changes need architecture team approval

How We Approach CI/CD

Continuous Integration:

All tests run on every commit
Builds are fast and reliable
Automated security scanning
Code quality checks (linting, formatting)

Continuous Deployment:

Automated deployment to staging
Manual approval for production
Feature flags for gradual rollouts
Automated rollback on critical failures

Monitoring:

Comprehensive observability
Real-time alerting
Performance tracking
User impact monitoring

Go-to-Market

How We Approach Go-to-Market

Our go-to-market strategy centers on building trust with the developer community and ecosystem partners.

Core GTM philosophy:

Developers first - Build for the people who actually use our products
Community-driven - Grow through genuine relationships and valuable contributions
Partner-enabled - Leverage ecosystem relationships for mutual benefit
Value-first - Lead with helpful content and tools, not just sales pitches

GTM channels:

Open source - Building trust through transparent, valuable contributions
Developer marketing - Technical content, tutorials, and tools
Ecosystem partnerships - Joint solutions with complementary providers
Conference speaking - Sharing expertise at industry events
Customer advocacy - Letting success stories speak for themselves

How We Keep Momentum

Content strategy:

Technical blog posts - Deep dives into networking, AI infrastructure, and distributed systems
Open source updates - Regular progress reports and feature announcements
Customer stories - How companies use Datum to solve real problems
Research reports - Original insights on AI infrastructure trends

Community engagement:

GitHub presence - Active in relevant open source communities
Discord/Slack communities - Participating in developer conversations
Industry forums - Contributing to standards and best practices discussions
Local meetups - Supporting developer communities in key markets

How We Think About Our Website

Our website should be the best resource for understanding modern network infrastructure for AI.

Website goals:

Educate - Help developers understand networking superpowers
Demonstrate - Show real examples of Datum in action
Convert - Make it easy to get started with our tools
Support - Provide comprehensive documentation and guides

Key pages:

Home - Clear value proposition and getting started path
Docs - Comprehensive technical documentation
Blog - Regular insights and updates
Community - How to engage with our ecosystem
Customers - Success stories and use cases

How We Think About Swag

Swag should be useful, high-quality, and reflect our engineering culture.

Good swag principles:

Actually useful - Things people will genuinely use
High quality - Reflects our commitment to excellence
Tasteful design - Professional but not boring
Limited quantities - Makes it feel special, not cheap

Swag ideas:

High-quality t-shirts and hoodies
Network topology posters or stickers
Developer tools (good cables, adapters)
Conference giveaways that start conversations

What Are the Most Common Use Cases

AI Infrastructure Providers:

Multi-cloud connectivity for inference workloads
Private connections between training and serving infrastructure
Global traffic routing for AI APIs
Observability across distributed AI systems

SaaS Companies Scaling Globally:

Private connectivity to customer VPCs
Performance optimization across regions
Compliance with data residency requirements
Cost optimization for egress traffic

Digital Enterprises:

Hybrid cloud networking
Partner ecosystem integration
Zero-trust network architecture
Application performance optimization

Modern Service Providers:

White-label infrastructure services
Customer private cloud connections
Multi-tenant network isolation
Automated provisioning and billing

What Is Our Brand’s Voice and Tone

Voice characteristics:

Technical but accessible - We know our stuff but explain it clearly
Confident but humble - We’re building something important but stay grounded
Direct but friendly - We don’t waste words but we’re human
Inclusive but expert - We welcome newcomers but maintain technical depth

Tone guidelines:

Use active voice and clear, simple sentences
Avoid jargon unless necessary (and define it when used)
Be specific with examples rather than abstract concepts
Show don’t tell - demonstrate value with real examples
Maintain optimism about the future we’re building

Writing principles:

Clarity over cleverness - Make it easy to understand
Useful over promotional - Lead with value, not just features
Authentic over polished - Be genuine about challenges and progress
Specific over general - Use concrete examples and metrics

Our Design Principles

Our brand personality and visual identity are guided by five core principles that reflect who we are and how we show up in the world:

Grounded and Visionary - We think big, but back it up with elegant, thoughtful products that solve authentic, real-world problems. Our designs balance ambitious thinking with practical execution.

Approachable and Expert - Our deep sector expertise is accompanied by an inclusive, collaborative, optimistic, curious ethos. We make complex technology accessible without dumbing it down.

Proactive and Driven - We constantly and transparently evolve in our quest to empower and support companies to build, invent, and compete, at scale. Our brand reflects forward momentum and continuous improvement.

Conscientious and Ethical - We always aim to do the right thing, whether helping businesses lower their carbon emissions, or treating all people with respect, dignity, and empathy. This shows through in how we communicate and what we prioritize.

Self Aware - We are humble enough to recognize when we’re not perfect, sound overly grandiose, or overdo the jargon, and aren’t afraid of poking fun at ourselves accordingly. Our brand has personality and doesn’t take itself too seriously.

Our Design Language

We maintain a foundational, hyperscale-ready framework that guides the design of all user experiences within the Datum ecosystem. This design language ensures consistency while allowing for the flexibility needed across different products and contexts.

Design philosophy:

Hyperscale-ready - Built to work at massive scale and complexity
Developer-first - Interfaces designed for people who think in code
Systematically consistent - Coherent experience across all touchpoints
Future-oriented - Designed for the AI-native world we’re building

Key design elements:

Typography - Clean, technical fonts that maintain readability at all scales
Color system - Professional palette that works across digital and physical applications
Iconography - Systematic icon library that speaks to our technical audience
Imagery - Photography and illustrations that show real technology in action

Brand applications:

Product interfaces and dashboards
Marketing websites and materials
Documentation and technical content
Conference presentations and swag
Social media and digital advertising

For detailed specifications, typography guidelines, color codes, logo usage, and design resources, visit design.datum.net.

This handbook is a living document that evolves with our company. Every team member can suggest improvements and updates.

Datum.net Company Handbook