Bridging Worlds
AI at the heart of every supply chain.
So operations can decide with the speed and clarity the market already demands.
What we're building
Amaru is an AI services platform for supply chain SMBs. Our technical core is inventory optimization through MILP (Mixed Integer Linear Programming) — mathematical models that calculate optimal stock levels, reorder points, and resource allocation under real constraints of capacity, budget, and demand.
Built for operations teams at small and mid-sized companies who today decide with spreadsheets and intuition, and need the same optimization power multinationals use — without the cost or complexity of a traditional ERP.
Optimal reorder points under real capacity & budget constraints
See it working
A working demo, from reading SAP data to a purchase plan. Screens below come from our internal prototype.
Illustrative demo with synthetic data. Demand comes from a public dataset (UCI Online Retail). Suppliers, costs and capacity are simulated. These are not client results, and the prototype has no connection to any live system.
Your purchasing data, read as it sits in SAP
The demo reads the tables a buyer already depends on: items, suppliers, stock by warehouse and sales history. Nothing is re-keyed and nothing is written back to the ERP.
A weekly purchase suggestion the buyer can review
Demand, variability and lead time per SKU become a reorder point, a safety stock and a suggested quantity. The buyer approves or changes every line.
An optimizer that plans across suppliers and weeks
A mixed-integer optimization model plans orders over the horizon, respecting lead times, minimum order quantities, supplier minimums and warehouse capacity, and shows the plan next to today's rule.
How to read: Order value by week
Each pair of bars shows how much money is ordered from suppliers in that week. Blue is the optimizer and orange is the current reorder-point rule. The rule orders in smaller amounts across most weeks, whenever stock crosses its reorder point. The optimizer groups purchases into fewer, larger orders, because each order to a supplier carries a fixed cost and has minimum-order requirements. The total spend over the horizon is similar. What changes is when and how the money is committed, and how many orders are placed.
How to read: Inventory of one product
This follows a single product (small popcorn holder) week by week. The dotted line is its safety stock. The orange line (current rule) reorders often, so stock zigzags around the safety stock. The blue line (optimizer) places one large order that arrives around week 5 and draws it down until the end of the horizon, so it carries more stock in the middle weeks and places fewer orders. It also lets stock dip briefly below the safety stock before that order arrives, because the model treats safety stock as a soft target with a small penalty. Both end the horizon at a similar level. Savings are judged on total cost across all products, not on one product alone. The holding, ordering and shortage costs used here are illustrative assumptions, so the shape of this curve would change with real client data.
Want to see it on your own data? We start with a diagnostic using the purchasing and inventory data you already have.
Start a conversationWho we serve
Companies with real purchasing volume — caught between spreadsheets and an ERP that's too heavy for what they need.
Between $10M and $50M — mid-to-upper-market SMBs, not micro-businesses.
A dedicated operations team managing real inventory and purchasing volume.
Deciding with spreadsheets, intuition, or legacy software — not mathematical optimization.
Traditional ERPs like SAP or Oracle are too expensive or too complex to implement at their scale.
How we work
Three steps, each one earning the next. You see results on your own data before you commit to anything long-term.
Diagnostic
We review your purchasing and inventory data, find where capital is tied up and where stockouts come from, and size the opportunity.
YOU GETA written diagnosis and a prioritized list of opportunities.
Pilot
We run the optimization model on your real data for a defined scope, such as one product family or supplier group, and compare its recommendations with your current decisions.
YOU GETMeasured results on your own numbers, not a generic demo.
Subscription
Once the pilot shows value, the model keeps running: updated reorder points, purchase recommendations, and regular reviews with your team.
YOU GETOngoing optimization and support as your operation changes.
We work from the data you already have. Scope and pricing depend on your SKU volume, and we share a proposal after an intro conversation.
Start a conversationYour data, handled with care
Purchasing data is sensitive. These are the commitments we make before you share any of it.
Read-only by design
Amaru proposes. Your team decides. We do not write to your ERP, and every recommended order can be reviewed and changed.
You control the data
We work with the extracts you choose to share, and we are happy to sign an NDA before you send anything.
Used only for your engagement
Your data is used to deliver your diagnostic. We do not use it to serve other customers, and we return or delete it when the engagement ends, on request.
Honest about our stage
We are an early-stage company and do not yet hold SOC 2 or ISO 27001 certification. We will complete your security questionnaire and agree controls that fit your requirements.
Frequently asked questions
What data do you need to get started?
For a diagnostic we typically ask for a purchasing and inventory extract: item master, current stock, purchase order history, supplier lead times, prices and minimum order quantities, and sales or demand history. A spreadsheet export is enough to begin. You choose what to share.
Do we need to connect to our ERP?
No. We start from exports, so there is nothing to install and no access to your ERP is required for the diagnostic. If the work continues, we scope any direct connection with your IT team. Our prototype demonstrates the approach on data structured like SAP Business One.
Does this replace our buyers and planners?
No. Amaru proposes a purchase plan and shows the reasoning and cost behind it. Your team reviews, edits and approves every line. Nothing is written back to your systems automatically.
How is this different from a reorder-point rule or a forecasting tool?
A reorder-point rule decides product by product. Our optimizer plans purchases across products, suppliers and weeks at the same time, within real constraints: lead times, minimum order quantities, supplier minimums and warehouse capacity. The result is a plan you can inspect, not a score from a black box.
What results should we expect?
We do not promise a number up front. The diagnostic sizes the opportunity on your own data, and you decide whether to go further. The savings shown in our demo are simulated on synthetic parameters and are not client results.
How long does it take and what does it cost?
Scope and pricing depend on your SKU volume, the number of suppliers and the quality of your data. We share a proposal with timeline and price after an intro conversation.
Who is Amaru built for?
US companies with roughly $10M to $50M in annual purchasing spend that buy from multiple suppliers and manage inventory in an ERP or spreadsheets, and that do not have a dedicated team of optimization specialists.
Do you handle perishable products?
Our current prototype covers non-perishable goods. Items with shelf life or spoilage need additional modelling, and we scope them case by case.
Another question? Tell us about your purchasing process and we will reply with how we would approach it.
Ask usWhere we're headed
In five years, a company's supply chain won't be managed between spreadsheets and 2 a.m. phone calls — it will be managed by a copilot that knows every supplier, every route, and every risk before it happens. Operations teams will stop spending hours reconciling scattered data across ERPs and portals; in minutes, they'll see where the bottleneck is, which supplier is failing, and what decision to make today.
A country's entire logistics will move with the precision of a system that learns from every exception and anticipates the next one. Small and mid-sized companies — today caught between giants running software built for another era — will access the same intelligence multinationals have, starting with inventory optimization via MILP as the first wedge.
Amaru will be the connective tissue behind that transformation: the quiet engine where data becomes action, and supply chain stops being a daily problem and becomes a competitive advantage you feel from the first deployment.
HIGHER TOGETHER