Artificial Intelligence

A useful app needs a reason for the next visit. Our development work connects plant identification to ongoing care, with design decisions you can inspect.

Maroš BednárRead more Mobile app development. Building Flori around plant care

I used to spend an hour a week piecing together search data from four tools. Now one overview Claude built from free sources does the job, and the new model cuts cache reads by 60%.

Maroš BednárRead more Claude Opus 5.5 costs a fifth less. Our web review takes 7 minutes

Shopify credits coding agents with changing the cost of native development. Its migration also shows the tooling and review work that any cost comparison must include.

Maroš BednárRead more Why Shopify is moving from React Native to native apps

What September AI releases mean for Slovak businesses and municipalities. Eurostat data, practical scenarios and a modelled calculation of time released.

Maroš BednárRead more September 2026 AI news and what it means for Slovakia

Three paragraphs produced 55,000 lines of game code and a score of 5.05 out of 10. The method behind it works on business deliverables too, and the honest numbers are the useful part.

Maroš BednárRead more The Gauntlet Loop. Give AI a Bar It Cannot Argue With

Measure the cost of a correctly completed task, including retries and human review. A verified price announcement and an illustrative calculation.

Maroš BednárRead more Measure AI Cost per Completed Task

Choose a first automation by scoring the work, data, risk, ownership, and reversibility. Then validate one pilot before you expand it.

Maroš BednárRead more What to automate first in a small business

Model origin is only part of the decision. Compare hosted APIs and self-hosting through data access, service terms and the cost of a completed business task.

Maroš BednárRead more Chinese vs American AI Models: Should Your Business Use Them?

A July 2026 comparison of GPT-5.6 Sol and Claude Fable 5. How to separate token prices from task costs and test vendor benchmarks against your own work.

Maroš BednárRead more GPT-5.6 vs Claude Fable 5: Which One Fits Which Job

API tokens can cost less than a developer hour, but tokens are not the whole project. A practical comparison of model usage, review, testing, and operating costs.

Maroš BednárRead more What AI-Assisted Development Costs in 2026: API Tokens vs Developer Time

MCP connects an agent to tools and data. A2A connects independent agents. Here is where each protocol fits and which security controls remain your responsibility.

Maroš BednárRead more MCP and A2A: How to Connect AI Agents to Business Systems

The AI Act is not only a legal topic. Companies need an inventory of AI tools, owners, data, risks, logs, human review, and vendor responsibilities.

Maroš BednárRead more AI Act 2026 Checklist for Chatbots, HR Tools, and Internal AI Workflows

An AI agent can save hours, but only if permissions, logs, approvals, owners, and failure paths are designed before it touches production systems.

Maroš BednárRead more AI Agent Governance Checklist Before You Connect CRM, ERP, or Email

Google says the basics still matter for AI Overviews and AI Mode. The difference is that weak content and technical debt have less room to hide.

Maroš BednárRead more AI Overviews and AI Mode: SEO for a Business Website in 2026

What should an adviser receive after a funding call is screened? An illustrative example separates an initial match from confirmed eligibility and shows how to assess a pilot.

Maroš BednárRead more Grant Screening for Advisory Teams

Connect documents, approvals and accounting. What to measure in a pilot and what MojaFirma demonstrates with synthetic data during active development.

Maroš BednárRead more AI Back Office System for Operations Teams

AI Act transparency rules apply from August 2, 2026. High-risk employment AI rules apply from December 2, 2027. Practical preparation guide.

Maroš BednárRead more Your AI hiring tool might be high-risk under the AI Act. Here's what that means.

An AI coding agent can speed up mechanical work. Production use needs clear checks, an accountable reviewer, and limited permissions.

Maroš BednárRead more AI in software development works best with a human in the loop

Dynamic Workers provide a place to execute generated code. What your team still needs to decide about permissions, network access and checking results.

Maroš BednárRead more Why your AI agents need a sandbox (and what Cloudflare just shipped)

Measure quality and cost per correctly completed task. Add routing when a cheaper candidate demonstrably handles a bounded share of the work.

Maroš BednárRead more Test the Case for a Multi-Model Strategy First

A model plan for one recurring task. Measure today’s work, set data permissions and output review, then decide whether the trial should continue.

Maroš BednárRead more How to Plan a 30-Day AI Pilot at Work

Everyone talks about AI agents. Most companies do not need one. Good automation is enough, and here is how to tell the difference.

Maroš BednárRead more AI agent or automation? When a Zap is enough and when you need an agent

From sorting messages to checking invoices. Ten illustrative workflows and practical checks for finding out whether AI saves time after review and corrections.

Maroš BednárRead more 10 Workflows for Testing Whether AI Saves Time

How to automate document processing, archive search and ERP integration. MojaFirma demonstrations, human review and a scoped pilot.

Maroš BednárRead more AI document management and archive search

The AI CyberLab proposal does not establish detection performance. Where to test AI in cyber defense and how to assess a pilot before deployment.

Maroš BednárRead more AI CyberLab: What's Realistic About AI in Cyber Defense Today

Correct language and a familiar voice do not establish identity. Verify bank-detail changes through a known contact and rehearse the team response.

Tatiana GavulováRead more AI Phishing and Deepfakes. Verify the Sensitive Request

GrantAI monitors grant calls, matches them to company profiles, prepares eligibility analysis, and drafts application materials. Demo and pilot for grant teams.

Maroš BednárRead more GrantAI prepares materials for advisers to review

AI helps me compare sources and challenge an argument. How I use it to prepare copy while keeping experience, claims and decisions with the author.

Tatiana GavulováRead more How I Use AI to Prepare Marketing Copy

Google describes protected cloud processing for selected Pixel features. See what the architecture protects and what the announcement does not establish for business data.

Maroš BednárRead more Google's Private AI Compute: Balancing Power and Privacy

Asking for several responses and their probabilities can help with creative tasks. What Verbalized Sampling studies, and why it does not make a clear brief obsolete.

Maroš BednárRead more Verbalized Sampling adds variety, not a replacement for a brief

Connect forms to CRM, assign enquiry owners and measure response time. Set up marketing sequences according to the contact purpose and sending rules.

Maroš BednárRead more Automated Leads from the Web by Connecting Forms, CRM, and Email Campaigns