The data centers are expanding, the funding rounds are closing, and anticipatory design is here. This week: infrastructure meets demand, startups solve real problems, and UX moves from reactive to predictive.
Industry moves
Centersquare drops $1B on data centers for AI workloads
Centersquare just acquired ten data center facilities across the US and Canada for $1 billion. The move brings their total portfolio to 80 data centers and reflects what everyone in AI infrastructure already knows: traditional computing facilities can't handle the power densities AI workloads demand.
The new facilities are spread across Boston, Minneapolis, Dallas, Tulsa, Nashville, Raleigh, Toronto, and Montreal. Two sites in Boston and Minneapolis were already under long-term lease with Centersquare. They just bought them outright.
Why it matters: AI doesn't run on hope. It runs on compute. And compute needs physical infrastructure that can handle density. As models scale and enterprises deploy more AI systems, data center capacity becomes the bottleneck. Centersquare's betting that consolidation and strategic expansion will capture the demand surge before competitors do.
"Infrastructure is unsexy until it's the constraint. Then it's everything. AI companies can build the smartest models in the world, but if they can't deploy them at scale, they're just expensive research projects. Centersquare's solving the unglamorous problem that actually matters."
{{Kyrylo Lazariev}}
Hipp Health raises $6.2M to fix behavioral health's admin crisis
Behavioral health clinics in the US are drowning. Autism and ADHD diagnoses are skyrocketing. Medicaid requirements keep shifting. Clinicians are spending more time on documentation and compliance than treating patients.
Hipp Health, a San Francisco startup, just raised $6.2 million in seed funding led by RTP Global to fix this. Their AI-native platform automates the operational work that keeps behavioral health practices running: clinical documentation, billing, compliance, scheduling.
The problem: Clinicians didn't get into medicine to fight paperwork. But that's where they spend their time. Hipp Health automates the unglamorous work so clinicians can focus on patients.
"This is product design at its core: remove friction from the workflow, automate the repetitive tasks, and let humans do what humans do best. The best interfaces are the ones users don't have to think about. Hipp's automating admin so clinicians can treat patients. That's solving the right problem."
{{Danylo Dubrovsky}}
5 European AI startups worth watching
Europe's AI ecosystem is maturing. These five companies are building real products, raising serious capital, and solving problems that matter.
- Black Forest Labs: AI-powered image generation and editing. Founded in 2024, already valued at $4 billion with partnerships at Meta, Adobe, and xAI. They're in talks to raise $300 million. For a one-year-old startup, that's aggressive momentum.
- Quantexa: Decision intelligence for fintech. Founded in 2016 in London, they closed a $175 million Series F in March 2025, hitting a $2.6 billion valuation. Their platform analyzes complex datasets to detect fraud, manage risk, and ensure compliance, all while respecting European privacy standards. Global banks and governments use their tech.
- Synthesia: AI video creation from text. Founded in 2017, now valued at over $2.1 billion. Over 90% of Fortune 100 companies use their platform to generate professional videos using AI avatars. No studios, no actors, no equipment. Just text-to-video with multilingual support.
- n8n: Workflow automation without extensive coding. Founded in 2019 in Berlin, n8n raised €55 million in March 2025 and is reportedly in talks for funding that could push their valuation above $1.5 billion. Over 200,000 active users connect 800+ apps using their visual, node-based editor.
- Tines: No-code automation for cybersecurity. Founded in 2018 in Dublin by former security engineers. They closed a $125 million Series C in March 2025, reaching unicorn status at $1.1 billion. Companies like GitHub, Coinbase, Canva, and Databricks use Tines to automate security alerts and incident response without writing code.
"What's notable here is the pattern: these companies are solving workflow problems. Automation, decision intelligence, content creation, security. All of them are removing friction from professional workflows. That's what drives adoption. Not flashy demos. Utility."
{{Oleksandr Koshytskyi}}
Product reality check
Meta launches business AI tools for SMBs
Meta rolled out Business AI tools accessible across its platforms and on businesses' own websites. They also unveiled augmented reality shopping experiences and customizable ad options.
The strategy: give small and medium businesses access to AI-powered tools that previously required enterprise budgets. Lower the barrier, scale adoption.
TikTok expands shop Ads support through Marketing Partners program
TikTok added a "Shop Ads" category to its approved Marketing Partners program — a directory of third-party providers that help brands manage their TikTok marketing.
Why it matters: TikTok's making it easier for brands to run shopping campaigns without needing internal expertise. More access, more adoption, more revenue.
Threads opens topic-based communities to all users
After testing with select users, Threads officially launched topic-based communities. Users can now find relevant discussions and like-minded people through curated topic feeds.
It's Threads' answer to discoverability. Twitter had hashtags. Reddit has subreddits. Threads is building topic-based communities to help users find conversations that matter to them.
Design wins
Lazarev.agency helped EllipX boost onboarding and streamline transactions
EllipX is a European crypto-finance platform combining a crypto exchange, digital wallet, and fiat payment operations under strict EU compliance. When they came to us, their product had just been redesigned, but the redesign failed to solve key UX challenges.
The problems:
- Users struggled with onboarding and verification.
- Navigation was confusing.
- Transaction flows caused drop-offs and abandoned actions.
- Incomplete verifications blocked business growth.
Our approach: We rebuilt the UX from the user's perspective. We simplified onboarding, clarified verification steps, streamlined navigation, and optimized transaction flows to reduce friction at every touchpoint.
The principle: Compliance-heavy products don't have to feel heavy. The challenge is balancing regulatory requirements with user experience. EllipX needed both. We delivered clarity without cutting corners.
Nike and Lego build a playground in Shanghai
Nike and Lego teamed up to create a playground at Baoshan No. 2 Central Primary School in Shanghai. The space features giant Lego blocks that kids can endlessly reconfigure.
Why it works: Both brands have distinct voices, but they complement each other. Nike brings movement and play. Lego brings creativity and modularity. Together, they built something kids actually want to use.
"Great collabs happen when both brands shine independently but amplify each other together. This isn't co-branding for press releases. It's functional, playful, and actually useful. That's rare."
{{Oleksandr Koshytskyi}}
Lazarev.agency client success updates
eDiscovery Assistant rebrands as Minerva26
eDiscovery Assistant, the legal research platform for litigation professionals navigating Electronically Stored Information (ESI), just rebranded as Minerva26.
The rebrand reflects their evolution from a knowledge platform into a strategic command center for discovery. The goal: equip litigation teams with foresight to lead, not react.
What changed: The platform now positions itself as a proactive tool, anticipating needs, providing clarity, and giving litigators control over ESI from the start.
"eDiscovery Assistant sounded like a helper. Minerva26 sounds like strategic infrastructure. The name change reflects the product's ambition."
{{Kyrylo Lazariev}}
Matta raises $5M to transform Africa's chemical supply chain
Matta, a digital marketplace for chemicals and materials, raised $5 million in a CRE Venture Capital-led seed round. The startup connects chemical producers with manufacturers and offers logistics and trade financing.
The opportunity: Africa's chemical supply chain is fragmented. Matta's building the digital infrastructure to streamline it, connecting suppliers, buyers, and logistics under one platform.
Events to monitor
- World AI Week (October 6–10, Amsterdam, Netherlands) Over 40 events covering AI in business, science, and tech. Topics include automation, ethics, creativity, and inclusion. If you're looking to connect with the global AI community, this is the week.
- MLOps / Gen AI World (October 8–9, Austin, TX, USA) For teams building or deploying large language models. Sessions cover LLM architecture, model evaluation, and cost-effective GenAI deployment.
This week's reality check
Why anticipatory design is the next big shift in AI-driven UX
Imagine your coffee starts brewing the moment you wake up. Curtains draw open as soon as your eyes open. Your map reroutes before you hit traffic. Once sci-fi. Now reality.
What is anticipatory design? It's designing interfaces that don't just respond to user actions, they predict needs and act before users request anything. The system anticipates based on user data and contextual signals: where you are now, where you're likely going next.
How is it different from personalization? Personalization reacts to preferences. Anticipatory design predicts intent and removes steps entirely.
The tech enabling it: Generative AI like GPT-4, real-time context modeling, and behavioral pattern recognition. These systems process inputs faster than users can articulate needs.
Where it's already live:
- Duolingo Max: Adjusts lesson difficulty based on predicted struggle points
- Netflix: Queues content before you search
- Intercom's Fin: Pre-loads support answers based on user behavior
- VTNews.ai: Surfaces news before readers know they want it
The design challenge: Anticipation requires trust. Get it right, and you're removing friction. Get it wrong, and you're creepy. The line between helpful and intrusive is thin.
"Anticipatory design is the ultimate test of UX maturity. You're making decisions on behalf of users before they ask. That requires deep behavioral understanding, technical precision, and restraint. The best anticipatory systems feel invisible, they remove friction without users noticing the system made a choice for them."
{{Oleksandr Holovko}}
What's coming next week
More infrastructure consolidation, more vertical AI plays, and hopefully fewer "AI-powered" features that are just API wrappers with no real intelligence.
🔥 Stay sharp. Stay with Lazarev.agency.
.webp)
