From Quote to Shipment,
Run and Improve Your Production

Optonum is a browser-based operations platform for manufacturers that runs the chain from quote to shipment in one place, laid out like the plant: Order to Delivery, Planning, Materials, Production, Quality, Maintenance, People and Improvement. Four AI agents propose changes worked out with formulas and the simulation; nothing changes until a person approves, and shift logs measure the result and show when the model needs calibrating. It sits beside your ERP and imports its spreadsheet exports, including those from Logo, Mikro and Netsis; it does not replace it.

First, plants of roughly 20 to 150 people that plan in spreadsheets beside an ERP and cannot say where the bottleneck is or whether a date is safe. Second, workshops of 1 to 20 people that do not want a heavy ERP. It is the wrong fit for a business that only needs accounting or e-invoicing, and it does not target semiconductors, power generation, aerospace and defence, or livestock and field farming.

Yes. The Free plan has no time limit and needs no credit card. It covers one plant, one project and two users, with the dashboard, the flow simulation, the OEE and takt calculators, AI Chat and AI Insights with daily limits, and five Bottleneck Oracle runs a month. Machines, lines, products and orders are unlimited. The paid plans (Workshop, Factory, Enterprise) add users, projects and agents, and are priced per organisation with you.

You see a first number on the day you sign up: the setup wizard asks up to four questions about the station that limits your output and shows the capacity those answers give. After that, data goes in this order — work centres, machines with cycle times, products with bills of materials and routings, production lines, then customers and orders. We have not yet measured how long a full setup takes at a real plant.

Any input can be left empty. A blank tile on the dashboard means the data behind it is missing; it never shows zero or an average dressed up as your number, and a rate you never entered is labelled assumed. The time-study, OEE and takt calculators work without plant data, so you can start with a stopwatch. What we cannot do is give a trustworthy answer from no data, and we say so rather than fill the gap.

Machines, bills of materials, routings, orders, quality readings and customers import from CSV or Excel, with Turkish column headers. Stock cards, customer lists, sales orders, open purchase orders and recipes exported from Logo, Mikro and Netsis import as they come out, recognised by their column names; there is no live connection to those three, and no mapping has yet been checked against a customer's own file. A read connector for SAP S/4HANA exists but has not yet been used with a customer's live system. General ledger, payroll and e-invoicing stay in your ERP or accounting package.

The product knows seven production types — job shop, discrete batch, repetitive, batch process, continuous, project and craft — and the type decides what the dashboard shows, which data it asks for and which advice makes sense. Job shop, discrete batch, repetitive and batch process are where most of what we have built and tested is concentrated; continuous, project and craft are supported and secondary for us. A line can carry its own type when a plant mixes two.

Turkish and English throughout: every screen, the help centre, and PDF and Excel exports. Each user chooses their own language, and the AI features answer in it. Numbers and dates follow the language (1.234,5 in Turkish, 1,234.5 in English), and the import reads either style.

No. AI Chat answers questions and changes nothing. AI Insights gives a briefing and changes nothing. The AI Command Center agents propose a change or guide the next step, and nothing is applied until a person approves it. The figures come from formulas and the simulation engine; the language model only writes the explanation.