36 months · forecast against actuals
Every forecast is graded against actuals each month, and we show you the gap.
The same staged structure as our actual scope-of-work documents. Each stage ends with a measurable result and a decision, and the engagement earns its next stage. The full page: how we work
Stage 1 · Scoping
We map your data, your systems (Priority, SAP, Hashavshevet), and the business decision the model must serve. The deliverable is a scope-of-work with success KPIs agreed before anything is built.
Every metric is defined against a baseline. For example: reducing 'purchasing error' by at least 10% from the current state.
Honest exit point. Characterization can conclude that AI is not the right tool yet, or that your data is not ready. The engagement can end here, and the findings are yours.
Stage 2 · Proof on real results
We build the capability on your real data and grade the results against the agreed KPIs. Not a demo: a measured test period with error tracking you can read yourself.
Honest exit point. If the results do not clear the bar we set together, the engagement ends here, with the full evaluation in hand.
Stage 3 · Staged rollout
Deployment into the systems you already run, as a staged rollout. Monthly grading against actuals continues in production, the same ledger discipline shown above.
You own what we build: models, code, and pipelines. No platform meter, no lock-in.
Multi-horizon sales and demand forecasts, graded openly against actuals every month.
Item-level forecasting, dynamic safety stock, and purchasing recommendations inside your ERP.
We would rather tell you now than in a kickoff meeting:
A three-year data-science engagement for a Mueller-group industrial manufacturer (US market). Forecasts graded against actuals every month; CEO recommendation on file.
For a security-sensitive organization: architected and specified a zero-egress AI enclave, with private networking end to end, an MLOps foundation, self-hosted LLMs, and vision object detection. Built and demonstrated a fully isolated, self-contained AI assistant: chat with organizational knowledge, retrieval over internal documents, deep-research workflows, document and diagram generation. Everything runs inside the closed network. Nothing leaves it.
Built and demonstrated a SAM-based system that automatically segments airfields in satellite and aerial imagery and measures them with precision tooling, wrapped in a purpose-built UI with manual correction, object addition, and report generation. It turns the costliest phase of geospatial vision work into a supervised, minutes-long task.
Dan Malka leads every engagement personally: the characterization, the build, and the handover. Whoever scopes your system here is the same person who builds it and integrates it, so the handover point where AI projects stall does not exist. The record is built from multi-year production engagements, not one-off deliveries. Dan replies within one business day.
We will tell you whether the data can carry it, what it would take, and how it will be graded.
