At AWS Summit Amsterdam 2026, I attended the Startup Theatre for session STP101: “AWS Loves Startups” — a fireside chat between AWS and Anthropic about how European startups can build and scale AI products on AWS infrastructure.
The Speakers
- Veronica Östring — Senior Startup Account Manager, AWS
- Alex King — GTM Cloud Partnerships, Anthropic
The session combined an overview of AWS startup programs with a fireside discussion about the Anthropic partnership and what it means for builders choosing Claude as their foundation model.
100+ Startups Building on AWS
The “How to grow your startup with AWS” slide showcased over 100 European startups already building on the platform. A few notable names across categories:
AI and ML:
- Mistral AI, PolyAI, Gladia, Hypatos, IRIS.AI, nancy.ai, Fuse
Security:
- Aikido, Hadrian, Upwind, Secfix, Tracebit
Fintech:
- Pleo, Wayflyer, Flatpay, Pair Finance, Suade
Developer Tools:
- Lovable, Storyblok, Depot, Cloudsmith, Tidio
Climate and Mobility:
- Tibber, tado, Monta, motorway
Infrastructure:
- Axiom, Hook, Torq, Xata, Zenity
This is a snapshot of the European startup ecosystem in 2026 — heavily skewed towards AI, security, and developer productivity.
AWS Activate: Credits to Extend Your Runway
Veronica Östring broke down the AWS Activate credit tiers:
- Founders: $1,000 in AWS credits — for self-funded, bootstrapped startups not affiliated with an Activate Provider
- Portfolio: Up to $100,000 in AWS credits — for startups funded up to Series A and affiliated with an Activate Provider
Key detail: AWS credits now cover third-party models on Amazon Bedrock — meaning your Activate credits pay for Claude, GPT, Mistral, Llama, and every other model in the catalog. That is a game-changer for AI startups managing inference costs.
The Unicorn Numbers
Two statistics that tell the story:
- 80% of startup unicorns run on AWS
- 96% of existing AI/ML unicorns run on AWS
When nearly every AI unicorn builds on your platform, that is not luck — it is ecosystem lock-in through developer experience, model availability, and infrastructure scale.
AWS Startup Program: Three Stages
The program follows a clear progression:
- Get Started — initial setup, Activate credits, technical guidance
- Build Fast — architecture reviews, Well-Architected Framework, solutions architects
- Accelerate Growth — co-selling through AWS Marketplace, enterprise customer introductions
Featured Activate Providers
The Portfolio tier ($100K credits) requires affiliation with an Activate Provider. The featured partners include:
- Y Combinator — the original startup accelerator
- Sequoia — legendary VC firm
- Andreessen Horowitz (a16z) — AI-focused mega-fund
- NVIDIA Inception Program — GPU ecosystem for AI startups
- Greylock — enterprise software specialists
- Carta — cap table and equity management
If your startup is backed by any of these (or 100+ other providers), you qualify for the full $100K in credits.
How AWS Activate Can Help: The Full Package
Beyond credits, AWS Activate provides a comprehensive support ecosystem:
- Up to $100K in AWS credits
- Up to $800K in exclusive offers and savings from partners
- Guidance from AWS experts and AWS Partners
- Community of 4,500 global ecosystem experts
- Showcase — extend your reach and visibility to enterprise buyers
- Cohort-based accelerator programs — structured growth tracks
- Training videos and learning content — technical upskilling
- Curated AWS templates and integration guides — production-ready architectures
The $800K in partner offers is the hidden gem — that includes discounts on tools like Datadog, Stripe, HubSpot, and dozens more that startups need beyond infrastructure.
Your AWS Startup Team
Once in the program, startups get access to a dedicated support structure:
- Account manager — your primary relationship contact
- Solutions architects — technical design and architecture reviews
- Investor managers — bridging startups with VCs and corporate buyers
- Specialist teams — deep expertise in AI/ML, security, containers, etc.
- Service teams — direct access to AWS product teams
- AWS Support — priority technical support
Startup Migrations Program
For startups migrating from on-premise, colocation, or other cloud providers, AWS offers a dedicated Startup Migrations program:
What you get:
- AWS Migration credits
- Guidance from AWS experts
- AWS Partner cash for hands-on migration support from vetted AWS Migration Partners
Featured migration partners: Cloud Nation, DoiT, CloudVisor, Cognitve, Automat-it, Netlight, cloudzone
The key insight: AWS will literally pay a partner to help you migrate off a competitor. That is how badly they want startup workloads.
Customer Story: Paebbl + Xebia — $250K Saved
“Big mission, small team — how Paebbl scaled with Xebia on AWS”
| Challenge | Approach | Outcome |
|---|---|---|
| Massive IoT data, no scalable platform | AWS Migration Program with Xebia | ”From idea to production in 2 hours” |
| Manual processes, weeks to deploy | Workshops → Landing Zone → real-time IoT pipeline | Deployment: weeks → hours |
| Needed solid foundations without hiring dozens | Xebia handles FinOps, GreenOps and managed services | $250,000 saved, 100% faster workflows |
“We needed solid foundations — without hiring dozens of engineers to do undifferentiated heavy lifting.” — David Pugh, VP Systems and Architecture, Paebbl
Customer Story: Legora — Legal AI with Claude on Bedrock
Legora — a collaborative legal AI platform serving hundreds of law firms and enterprises globally:
- Goal: Find the best performing foundation models for their individual features
- Approach: Partnered with AWS and Anthropic to do model evaluations and prompt migrations; implemented Claude Opus 4.6 on Amazon Bedrock
- Result: Amazon Bedrock now powers Legora’s collaborative legal AI platform; ready for next GenAI capabilities with Claude Sonnet 4.7
This is a perfect example of the AWS + Anthropic joint support: helping startups evaluate, migrate, and optimise their model stack.
AWS Partner Network: Scale Through Co-Sell
For startups ready to sell through AWS, the Partner Network unlocks serious growth levers:
- Access to AWS POC credits and Marketing Development Funds
- APN partner status and competencies — demonstrate expertise and credibility
- Architectural optimization for performance, security, and scalability
- Joint marketing, sales enablement, and co-selling initiatives
- Increase visibility and reach new customers with AWS Sellers
- Opportunities for partnerships, integrations, and business development
Co-Sell: Join the AWS Partner Network
The AWS Partner Network (APN) is a global community leveraging AWS technologies, programs, expertise, and tools to build solutions for customers. Key stats:
- 140,000+ partners from over 198 countries
- 70% headquartered outside the United States
Go-to-Market: AWS Marketplace
A digital catalog that makes it easy to offer software, data, and other products to AWS customers:
- 20,000+ product listings
- 5,000+ sellers
- 2.5M+ subscriptions
- Deployed in 30 regions
- 70+ categories (security, databases, generative AI, etc.)
- 300,000+ active subscribers
Customer Story: Aikido — “Secure Everything. Compromise Nothing.”
Aikido — a security startup that went all-in on the AWS partner GTM motion:
| Co-Build | Co-Market | Co-Sell |
|---|---|---|
| AWS ISV Software Partner certification | AWS Marketplace Seller Prime Program | 20% more customer enquiries within weeks of Marketplace launch |
| Native AWS platform integration | Paid search campaigns on AWS.com | 30-day time to revenue: first wins within first partner mapping cycle |
| Marketplace Private Offers + Self-Service | Co-branded content placement | |
| Best Practice: Self-managed onboarding | Best Practice: Phased rollout (Enterprise → Public) |
How Selling on AWS Marketplace Helps You Scale
The numbers speak for themselves:
- 80% richer deal sizes — strengthen co-sell relationship with AWS account teams
- 27% higher win rates — sell the way your customers want to buy
- 40% faster sales cycles — improve efficiency and profitability
Resources: Prompt Library and Workshops
Veronica Östring closed with practical resources for builders:
- Prompt and Agent Library — pre-built prompts curated by AWS experts
- startups.aws — the central hub for all startup programs
- workshops.aws — hands-on implementation guides and templates
Fireside Chat: AWS Startups x Anthropic
The session closed with a fireside chat between Veronica Östring (Senior Startup Account Manager, AWS) and Alex King (GTM Cloud Partnerships, Anthropic).
Key themes from the conversation:
- How AWS and Anthropic jointly support startups building on Claude via Amazon Bedrock
- The model evaluation and prompt migration process (as demonstrated by the Legora case study)
- Best practices for choosing between Claude model tiers (Haiku for speed, Sonnet for balance, Opus for complex reasoning)
- The roadmap from prototype to production-scale AI deployment
This partnership gives startups direct access to both AWS infrastructure expertise and Anthropic model optimization guidance — a rare combination in the current AI landscape.
Get Started
The session closed with a simple CTA: startups.aws — home to over 50,000 startups globally.
The AWS + Anthropic Partnership
The fireside chat revealed how deep the AWS-Anthropic relationship goes beyond just hosting Claude on Bedrock:
- Joint GTM programs — helping startups go from prototype to production with Claude
- Credits and support — AWS Activate provides infrastructure credits while Anthropic provides model access
- Technical collaboration — optimising Claude inference on AWS custom silicon (Trainium, Inferentia)
- Enterprise bridge — startups building on Claude via Bedrock can scale to enterprise customers through AWS Marketplace
This aligns with what we saw in the keynote: AWS announced Bedrock Managed Agents powered by OpenAI (including GPT 5.5), but Anthropic remains the strategic AI partner with the deepest integration across the AWS ecosystem.
My Take
For startups evaluating their AI stack in 2026, the message is clear: pick your model (Claude, GPT, Mistral, Llama), deploy on Bedrock for the operational benefits (security, guardrails, observability), and leverage AWS startup programs for the credits and go-to-market support.
The competitive landscape between AWS, Azure, and GCP for startup mindshare is fierce — and having Anthropic on stage is AWS’s strongest card in that game.
I attended AWS Summit Amsterdam 2026 on May 27. Read my full keynote coverage for the Bedrock AgentCore and KBC announcements. For AI platform strategy, book a consultation.
