Forecio BUILD · AI development

We are a development team and a startup house. We build software for clients and found and run projects of our own, from the GEIM sportswear brand to the Lignufy marketplace. AI shortens our development time, but we know what to build from running our own businesses and from dozens of CRM and ERP projects for clients. We build your tool, copilot or full product and connect it to the systems you already use.

Discuss your project or call us +420 355 444 113
Own projects: GEIM, WAVE, Lignufy and more.
Examples from our products

WAVE: weekly report sent to Slack

Select an area to see what happens there

Illustration, not real data

Companies that trust us:

  • GoodData
  • Yieldigo
  • Shoptet
  • Chromozoom
  • Whalebone
  • Remax
  • Jungheinrich
  • Investown
  • Safetica

What we build on

  • 5+ projects of our own that we built and launched GEIM, WAVE, Lignufy, HledejLéky, Sales Copilot and more
  • 150+ clients in 12 countries CRM, ERP and integration projects
  • Yours code and data we hand over source code and access

AI speeds up writing code. People who know the business and its operations decide what to build.

No demos. This is what we have built.

We build for clients and for ourselves. We built and launched every one of these, and more keep coming.

Lignufy: a B2B timber marketplace

Timber sellers and buyers across the Czech Republic exchange offers and requests, place orders, chat and get notifications. We connected the platform to HubSpot and built its go-to-market around it, from the website to outreach across seven customer segments.

Connected to: HubSpot API, email. Status: live, more modules planned.

lignufy.com
Lignufy: a B2B timber marketplace

WAVE: a company dashboard instead of an ERP

A young D2C brand on Shopify had no ERP and no single view of its numbers. The dashboard combines the online store, analytics, ads, email and stock. Every Monday an AI report lands in Slack with an assessment of the week, unit economics, a stock forecast with a reorder date and concrete recommendations.

Connected to: Shopify, GA4, Meta, Google Ads, Sklik, Klaviyo, Clarity, Slack. Status: live, EDI integration with retailers is next.

WAVE: a company dashboard instead of an ERP

GEIM: a sportswear brand from scratch

A Czech functional clothing brand made from organic materials with a graphene coating. With the GEIM team we built the online store and the market launch, from strategy to social media content that we prepare with the help of AI.

Status: online store live since the end of 2025.

GEIM: a sportswear brand from scratch

Sales Copilot: a sales assistant

It turns buying signals into an account strategy, finds the right people, writes a five-step sequence and, once approved, logs everything in HubSpot. Every morning it summarises what happened and keeps track of deadlines, deals and follow-ups. A person approves every step.

Connected to: HubSpot, Amplemarket, Gmail, Slack. Status: in use internally, client version in preparation.

Sales Copilot: a sales assistant

HledejLéky: a medicine price search engine

A search engine that compares prices of over-the-counter medicines across online pharmacies and links to the cheapest one. We built the search, pharmacy profiles and a partner model.

HledejLéky: a medicine price search engine

Integration middleware

Our own middleware connects CRM with ERP, online stores and data warehouses, with monitoring and automatic retries. Clients including Slevomat rely on it.

Connected to: HubSpot, ERP, data warehouse via API. Status: live at clients.

Integration middleware
Integration middleware

How we work with AI

Where AI works and where people decide

The code, data and access stay with you.

  1. Specification and prototype

    AI turns the workshop into a specification and a clickable prototype within days. You approve the scope and priorities with our consultant.

  2. Development

    AI writes the first version of the code, tests and documentation. A developer designs the architecture and reviews every change before release.

  3. Data and integrations

    AI speeds up data mapping and migration. We test every connection on real data in your systems.

  4. Operations

    AI monitors for errors and suggests fixes. A person decides what comes next, based on how the tool performs.

What we build

Five types of solutions, one team from analysis to production

We start from the process, not the technology. If an off-the-shelf tool does the job, we recommend it.

Internal tools and dashboards

Dashboards, reporting, approvals and record-keeping where spreadsheets fall short and a full ERP is overkill.

  • One view of numbers from several systems
  • Automatic reports to Slack or email
  • Roles, permissions and an audit trail
Discuss your project (Internal tools and dashboards)

How we work together

From idea to first version

You get a time range after the discovery sprint. An internal tool or copilot usually takes weeks, a product more often months.

  1. Discovery sprint

    1 to 2 weeks: goals, processes, scope, a clickable prototype and an estimate as a range. You leave with a decision, not a slide deck.

  2. First version

    Weekly iterations, with a demo on real data every week.

  3. Integration and launch

    Connections to your CRM, ERP and other systems, team training and usage tracking.

  4. Operations and growth

    Support, SLA and new features based on how the tool performs.

Have an idea, or a process that slows your team down?

Free project consultation

In 45 minutes with a developer and a consultant, we go through your brief together and you get a short summary. We tell you honestly whether custom development makes sense or an off-the-shelf tool will do. No obligation.

Book a consultation

Mainly need to connect systems? CRM and ERP integration

  • Goal, users and the process the solution supports
  • Custom build or off-the-shelf tool
  • Connections to your systems and data
  • Next step and a rough time range

Talk to our team

Let's discuss your project

  • Lukáš Lískovec Lukáš LískovecFounder and CEO
  • Tomáš Drvota Tomáš DrvotaHead of Development
  • What happens: An online call with a developer and a consultant about your brief.
  • What to prepare: A short description of the problem, who will use the tool and which systems you use today.
  • What you get: A recommended next step, for example a discovery sprint. No obligation.

Prefer to write? Leave us your email and we will get back to you.

Book a call with our team

Prefer to call? +420 355 444 113 · Open in new window (opens in a new window)

Book a call Call

Frequently asked questions

Both. We build custom software for clients, and we found and run projects of our own, such as GEIM, WAVE and Lignufy. That is how we know what building and running a product really takes.

An internal tool or copilot usually takes weeks, a product such as a marketplace more often months. You get a firm range after the discovery sprint.

It depends on the scope, the number of connected systems and how much an off-the-shelf tool can cover. After the consultation and discovery sprint you get the scope of the first version and an estimate as a range, not a shot in the dark.

You do. We hand over the source code, data and access, so you are not dependent on us.

A developer reviews every change before release, and the code goes through tests and security checks. AI speeds up writing code, while people decide on architecture and releases.

Yes, connecting systems is our core work. We work with HubSpot, Salesforce, Pipedrive, Zoho, Odoo and clients' own systems.

When the process gives you a competitive edge, off-the-shelf tools cannot handle it without workarounds, or licence fees would cost more than building your own. Otherwise we recommend the off-the-shelf tool.

We choose by task, cost and data requirements. For AI agents we mainly use Claude models, connected to your tools through APIs and MCP (Model Context Protocol). We pick the application stack so that your own team can take it over.

We do, through support and ongoing development, or we hand it over to your team with documentation.