Claude Certified Architect — Limited Spots
We help companies design, build, and deploy intelligent AI agents using the full Claude ecosystem — from strategy and team training to production-ready implementation with Claude API, Claude Code, and MCP.
Free · No obligation · Reply within 1 business day
Or directly: info@blocksize.hr +385 99 383 4923
Process supplier invoice #4471 and match it against the purchase order.
I need the purchase order from the ERP, the invoice line items and the approval rules.
erp.get_purchase_order(4471)
→ Purchase order PO-4471 · 3 line items · €12,400.00
documents.extract("invoice-4471.pdf")
→ 3 line items · €12,400.00 · due in 30 days
Line items and amounts match (3/3). Within the auto-approval threshold.
erp.post_invoice("PO-4471")
→ Posted · payment scheduled for Oct 18, 2026
Invoice processed in 41 s. No human intervention needed.
Illustrative demo of an agent using MCP tools. The scenarios are fictional.
Reduction in processing time for clients
Faster onboarding of new employees
Automation of repetitive tasks
From inquiry to tailored AI plan
Services
From first strategy session to deployed AI agents — we cover every stage of your company's Claude adoption journey.
We identify where Claude AI creates the most value in your business and build a roadmap to get there.
We build production-grade AI agents and integrations using the full Anthropic toolkit — Claude API, Claude Code, MCP, and the Agent SDK.
Upskill your entire workforce — from C-suite to engineering — with certified Anthropic training programs.
How We Deliver
Every engagement is tailored to your company's size, goals, and technical maturity — no cookie-cutter programs.
Quick Start
Half-day
Identify your highest-ROI Claude use cases
Team Training
Full-day
Your team builds with Claude, live
Enterprise
Multi-week
Strategy to production deployment
Ongoing
Monthly
Continuous optimization & support
Expert Insights
Articles on Claude AI strategy, agent implementation, the MCP protocol, and trends in business AI adoption.
Can you let managers ask the database in plain language? We read the BIRD and Spider 2.0 leaderboards as of 29 October 2026: 83.4% vs 93% for human experts, enterprise tasks from 17% to 97%, why the agent around the model matters more than the model, SQL dialect gaps and a checklist for production.
AI crawlers now read the web far more than they send readers back. We analysed Cloudflare Radar data for October 2026: crawl-to-refer ratios of OpenAI, Anthropic, Perplexity, Google and Mistral, the split between training, search and user fetches, the most blocked bots in robots.txt and a practical policy for company websites.
Before an AI application scales, it hits a rate limit. We compared the official 2026 limits of Anthropic, OpenAI, Google Gemini, Mistral and Manus: usage tiers, tokens per minute, monthly spend caps, the two kinds of 429 error and why prompt caching can raise Claude throughput fivefold.
Contact
Tell us about your company and goals — we'll send you a tailored proposal showing exactly how Claude AI can impact your bottom line.
You'll hear from us within one business day with a tailored AI roadmap for your company.