End-to-End Audit and Implementation of an IT System for Optimizing the Logistics Processes

1. Reduction of logistics costs (fuel, mileage, downtime)

   • Metrics: liters per trip, ₸/km, percentage of optimization relative to baseline.

2. Increased delivery productivity without expanding staff

   • Goal: Achieve higher delivery output using the current fleet and personnel (“reducing future drivers who have not yet been hired”).

3. Reduction of undelivered orders and closure of open weekend routes

   • Metrics: number/percentage of undelivered orders by day of the week; lost revenue (₸) vs. baseline.

4. Cleaning and control of the Active Client Base (ACB)

   • Metrics: share of verified coordinates.

5. Transition from manual operations to status-based management

   • Metrics: order picking/loading time, percentage of orders with complete and correct statuses, number of manual route adjustments.

6. Integration with 1C and establishment of stable data exchange

   • Metrics: completeness and quality of data exports, data exchange SLA, integration stability in pilot and production environments.

7. Training and operational readines
   • Metrics: percentage of trained staff, UAT results, and production launch readiness criteria.

• Operational Efficiency:

  Reduction in total mileage and fuel consumption, faster generation of optimal delivery routes, and decreased manual intervention in route management.

• Service and Revenue Impact:

  Reduction of under-deliveries, improved on-time delivery performance, increase in average order value through better ACB point management, and recovery of previously lost weekend revenue.

• Transparency and Manageability:

  End-to-end delivery status tracking, unified analytics across sales and logistics, clearly defined roles and responsibilities, and regular structured reporting.

· logistics management platform

· Microsoft 365 (Teams, Outlook, Excel for reporting)

· 1C:Enterprise (ERP integration)

· Windows Server (RDP access)

· Google Maps API (for geocoding and route optimization)

• Legacy ERP and integration.

  The project required establishing access (RDP to the test database), agreeing on technical specifications, and ensuring correct mapping of directories, events, and statuses between 1C and Logistic system. It also required taking into account the limitations of legacy data structures and exchange performance.

  Data quality and geocoding.

• The process involved cleaning ACB addresses, assigning geocoordinates, and manually correcting “undefined” addresses together with the client’s internal team. Without this step, routing algorithms would lose accuracy and efficiency.

• Algorithm configuration.

  The tuning of routing algorithms included parameters such as delivery time windows, priorities, traffic and road constraints, and minimum order thresholds. Additionally, delivery zones and weekend routing scenarios were reviewed and adjusted on a semi-annual basis according to demand, distance, and load metrics.

Organizational Complexity

• Multifunctional coordination.

  The project involved multiple departments — logistics, order desk, warehouse, sales, and IT — as well as multiple parties (Client, Developer, Consultant). This required a clear organizational structure, well-defined communication channels, and decision protocols. These elements were formalized within the project role model and supported by weekly status meetings.

• Shift schedules and continuous production.

  Transitioning to status-based operations and new logistics procedures had to be done without interrupting daily deliveries. This was achieved through role-based training, a gradual introduction of new control mechanisms, and a staged rollout plan.

• Behavioral change management.

  Replacing manual route composition and reassignment practices required new levels of process discipline and KPI transparency. These behavioral shifts were supported by analytics dashboards, structured progress tracking, and regular communication across all stakeholder groups.

Timeline Constraints and Risks

The project operated under tight timelines for testing, integration, and training. It depended on data readiness and the workload of the client’s IT team, as well as external factors such as seasonality and regulatory constraints.

The Relog platform was implemented to eliminate key bottlenecks: inefficient routing, manual order handling, weekend under-deliveries, and gaps in Active Client Base (ACB) management.

Phasing and Workflow

• Diagnostics (Process Audit).

  Comprehensive analysis of current workflows, including route planning, order dispatch, warehouse coordination, and delivery management. The audit identified excessive manual work, lack of delivery status tracking, and significant dependence on human decisions. The output was a detailed AS-IS model and a TO-BE vision with automation, route optimization, and integrated status control.

• Technology Scouting and Selection.

  Identification and assessment of relevant technologies applicable to logistics optimization — including AI-based routing, data-driven dispatching, and ERP-integrated analytics.

  Evaluation criteria covered scalability, interoperability with existing 1C infrastructure, vendor maturity, cost of ownership, and ability to support regulatory requirements specific to the Kazakh food industry.

• IT Solution Analysis and Fit Assessment.

  Benchmarking multiple IT solutions against the identified technology stack. Comparison was based on functional coverage, integration potential, API compatibility, total cost of ownership, implementation speed, and vendor support.

  As a result, was selected ahe optimal platform, combining advanced routing algorithms, geocoding capabilities, and seamless integration with 1C:Enterprise ERP.

• Strategic Option Comparison and ROI Evaluation.

  Several implementation scenarios were compared — from in-house development to off-the-shelf adoption and hybrid models. Each was analyzed across cost-benefit, payback period, organizational change impact, scalability, and risk.

  The selected platform demonstrated the highest ROI potential, supported by measurable efficiency gains, savings in fuel, reduction in under-deliveries and low implementation risk due to modular architecture.

• Integration Framework.

  Setup of RDP access to the test 1C database, approval of integration specifications (dictionaries, orders, statuses, events), and phased testing of data exchange between systems.

• Data Cleansing and Geocoding.

  Address normalization, coordinate assignment, and mapping of the ACB to the full city/region geography. Filters by client profile (network vs. independent stores, average check, HoReCa indicators) were configured to support route prioritization and sales analytics.

• Routing and Zone Balancing.

  Activation of Relog’s optimization algorithms for route sequencing based on time windows, traffic, and road constraints. Zones are reviewed semi-annually based on changes in order density, distance, weight, and delivery volume.

• Weekend Order Management.

  Redistribution of orders to close “open” routes on weekends, minimizing under-deliveries in low-priority areas and ensuring minimum order thresholds are met.

• Commercial Analytics Layer.

  Visualization of active and inactive clients, revenue segmentation, and identification of high- and low-performing delivery zones. Joint analysis with the sales department helps increase order frequency and the average order value per client.

• Project Governance and Training.

  Weekly coordination meetings with all stakeholders, continuous plan tracking, issue resolution logs, and role-based user training to prepare for the go-live phase.

Kazakhstan — City of Almaty, with potential scalability across regional distribution centers.