USD–KZT Forecasting for Mining — Dynamics 365 partner Awara IT

Kostanay Minerals faced material earnings volatility driven by fluctuations in the USD to KZT exchange rate. Manual forecasting using legacy tools required a week of analyst effort and produced inconsistent accuracy, which slowed financial planning and increased exposure to FX risk.

Awara IT implemented an Azure Machine Learning predictive model integrated with Power BI Premium, trained Kostanay’s analytics team, and operationalized the model as a scheduled service. The solution reduced forecast preparation from one week to one hour, cut estimated IT costs by $35,000, and delivered repeatable, auditable forecasts that can feed into financial planning and hedging workflows. Awara IT served as the Microsoft Dynamics 365-capable implementation partner, ensuring the design is integration-ready with Dynamics 365 Finance for downstream treasury and planning processes.

Client / Industry / Country

Kostanay Minerals is a Kazakhstan-based mining company focused on extraction and processing of mineral commodities. A mid-sized regional operator, the company runs several sites in the Kostanay region and maintains centralized finance and analytics teams that support operational decision-making across the business.

Business challenge

Exchange-rate volatility between the US dollar and the Kazakhstani tenge (KZT) had a direct and measurable impact on Kostanay Minerals’ margins and cash-flow forecasts. Existing forecasting relied on spreadsheets, Tableau extracts and 1C export pipelines that required manual reconciliation, producing forecasts that took up to a week to prepare and validate.

The company also lacked an automated retraining and deployment workflow and had limited in-house experience with cloud machine learning. This combination increased the cost of maintaining forecasts, slowed the company’s ability to respond to FX moves, and constrained the finance team’s ability to integrate forecasts into timely hedging or operational decisions.

Why Dynamics 365 was selected

The engagement used Microsoft Azure Machine Learning and Power BI because the Microsoft stack provided the most straightforward path to operationalize predictive models and to integrate results with enterprise finance systems. Key selection criteria were security and identity integration (Azure Entra ID), ease of integration with Power BI Premium for reporting, and enterprise-grade model management available in Azure ML.

Awara IT was engaged as the implementation partner because of its Microsoft ecosystem expertise, including Dynamics 365 capabilities. Although the project focused on Azure AI and reporting, Awara IT designed the solution to be Dynamics 365 Finance–ready so forecast outputs can be consumed by financial planning, treasury and hedging modules in Dynamics 365 with minimal additional integration work.

Modules implemented

• Azure Machine Learning: predictive model development, training and model registry
• Power BI Premium: interactive dashboards and scheduled report refresh
• Azure Blob Storage: centralized historical and feature store
• Data pipelines: ETL/ELT for market feeds and legacy 1C extracts
• Model operations: automated retraining, versioning and monitoring
• Training & knowledge transfer: upskilling Kostanay’s analytics team on Azure ML

Integrations

• Power BI Premium (reporting and visualization)
• Azure Blob Storage (historical FX and feature storage)
• Legacy 1C exports (financial history and operational inputs)
• External FX market feeds (pricing and macro indicators)
• Azure Entra ID (authentication and role-based access control)

Localization & compliance

The solution was configured for Kazakhstan-specific requirements: KZT currency formatting, local timezone considerations, and Russian/Kazakh-friendly report labels where required. Data residency and protection were handled by applying encryption at rest and in transit, role-based access via Azure Entra ID, and retention policies aligned with Kostanay Minerals’ IT governance.

Reports and audit logs were structured to support local regulatory reporting needs and internal controls so outputs could be validated by finance and compliance teams before being used in statutory or tax-related submissions.

Business value

Kostanay Minerals gained a repeatable, auditable forecasting capability that materially reduced the operational burden on analysts while improving the quality and timeliness of FX information used in planning and treasury. The model’s integration with Power BI accelerated insight consumption across finance and operations teams and created a single source of truth for USD–KZT projections.

Beyond immediate time and cost savings, the engagement delivered upskilling for the internal analytics team and a production-ready deployment pattern. With Awara IT’s Dynamics 365-aware design, forecast outputs can be routed into Dynamics 365 Finance to operationalize hedging strategies and align treasury workflows with corporate planning.