Process audit & mapping
We trace how requests, money and documents actually move. You get a process map and an automation plan with priorities.
Automation · AI · ERP
Business automation and AI adoption
We bring order to operations: records move from spreadsheets into a system with roles and history, AI assistants join the team, and a custom ERP grows around your processes — all on infrastructure you control.
01What we do
From process audit to your own ERP
We do not sell a box. We study how your business actually works and build exactly the system it needs.
Process automation
Records and approvals move from spreadsheets and chats into a system: roles and permissions, statuses, reminders, full change history.
AI adoption
Assistants working on your company's data: executive summaries, document processing, answers to your team's questions.
Custom ERP · Integrations · Infrastructure & support
Custom ERP
A management system built for your company: only the modules you need, your terms and your processes — no paying for “everything at once”.
Integrations
We connect what already works: 1C, Google Workspace, Telegram, payment services — via API, with no double entry.
Infrastructure & support
Deployed on your server or cloud: monitoring, backups, alerts. The system lives on — and someone is accountable for it.
02what it looks like
Four screens from an ordinary working day
Not mockups and not screenshots: these interfaces are built the same way we build them for clients. The data is illustrative.
sample data
New5
Equipment replacement, shop 2
Customer return
Request #1284
In progress7
Packaging delivery
Forklift repair
Done41
Supplier agreement
One request going through approval
Eight steps and no trace
Every step lives in its own channel, and nobody sees the whole picture.
- 01An employee posts the request in a chat
- 02The manager says "fine" out loud in a meeting
- 03Accounting asks for a copy by email
- 04The file is edited in two copies at once
- 05The finance director is away — everyone waits
- 06Nobody remembers who approved it
- 07The payment goes out late
- 08A month later someone hunts for where it got lost
The timeline is luck. The owner is whoever agrees to be.
What this looks like across industries
typical scenario
Retail chain
before
Nine stores, each with its own stock spreadsheet. The summary was compiled by hand on Mondays and was stale by Wednesday.
what we did
Unified stock and sales accounting, automatic purchase requests at minimum levels, a live report for the owner.
after
Stockouts stopped being a discovery. Orders are generated by the system, not by someone's memory.
Summary: from two days to seconds
typical scenario
Logistics
before
Requests arrived by email, chat and phone. Only the dispatcher handling a run knew its status.
what we did
A single request desk with statuses and owners, integration with 1C and Telegram, full change history per request.
after
Anyone can see where the cargo is and who is responsible. A dispatcher's holiday no longer stops the work.
Lost requests: none
typical scenario
Manufacturing
before
The output plan lived in Excel, the actuals in the foreman's logbook. Discrepancies surfaced at month end.
what we did
Output recorded per shift, material consumption norms, automatic batch cost calculation.
after
Overconsumption is visible the same day, not in a report a month later.
Batch cost known on the day of production
typical scenario
Services
before
Contracts sat in folders and renewal dates in the lawyer's head. Approvals went by email and left no trace.
what we did
A contract register with statuses and reminders, approval routes, access rights and complete history.
after
Not a single missed renewal. It is visible who holds a document and for how many days.
Approval: from 11 days to 4
We run our own business on this platform
Our best case is PMEG's own ecosystem: we manage real construction projects on these tools every day. Below are working products, not mockups.
Executive dashboard · Field data system · AI analyst · Control center
Executive dashboard
The whole project on one screen: budget, earned value, forecast, critical path. Opens on any device.
Field data system
A replacement for working spreadsheets: role-based data entry, month locking, and a log showing who changed what and when.
AI analyst
Answers questions about project data and prepares executive briefings: what is late, where the budget goes, what needs a decision.
Control center
The whole infrastructure on one screen: services, databases, backups, disk. Problems are visible before anyone reports them.
03builder
Assemble your own solution
Tick what hurts. We will show the make-up of the system and the time to a working prototype.
what hurts today
what we would build
Tick at least one item on the left and we will assemble the scope and the timeline.
04How we work
A working prototype first, the big system second
We do not start with a hundred-page spec. You decide on the big system while looking at a working screen with your own data.
01 · 1–2 weeks
Audit
We study how processes and records work today and where time and money leak.
02 · 2–4 weeks
Prototype
A working system on your real data — something to touch, not a slide deck.
03 · from 1 month
Rollout
We turn it into a product: permissions, integrations, team training. Launched in stages, without stopping the business.
04 · ongoing
Support
Monitoring, backups, new features as you grow. The system grows with the business.
05FAQ
Frequently asked questions
How is this different from an off-the-shelf CRM or 1C?
An off-the-shelf system makes the business adapt to it. We do the opposite: the system mirrors your processes and your vocabulary, has no unnecessary modules — and you do not pay for what you do not use.
Everything we have is in Excel. Is that bad?
Up to a certain size it is fine. It becomes bad when spreadsheets start costing money: lost edits, double entry, manual reports. We migrate data carefully and usually run the system in parallel with the sheets until the team settles in.
How much does it cost?
It depends on scope, which is why we start with an audit and a prototype: a limited budget and a tangible result, after which the cost of the full system can be estimated precisely.
What about confidentiality with AI?
The assistant only works with the data you grant it, on infrastructure under your control. What is shared and where is fixed in writing before we start.
Who maintains the system afterwards?
We do: support is part of the service, from monitoring and backups to new features. The code and documentation are yours — we can hand the system over to your team if you wish.
06Contact
What would you like to automate?
Tell us about your process and current tools. We will discuss the first phase and show a relevant demo.