Let's get in touch and discuss your case
Let's start negotiating and develop your own unique project. Our team is ready for the most daring projects.
Route‑to‑Market Optimisation for Medical Representatives
Together with Taal Healthtech, we reshaped not just the routing algorithm, but the way field teams operate: who to visit, in what sequence, and how to spend the working day for maximum impact.
- Used materials: historical GPS traces, CRM visit data, mapping and live‑traffic APIs
- Industry: HealthTech
- Time spent on development: 3 months
01.
Client's Request and Solution
The client sought to increase field efficiency by optimising routes and visit prioritisation, embedding this into day‑to‑day operations via tight CRM integration and moving to KPI‑driven planning. The practical challenges were familiar: data quality issues (unstandardised addresses, missing or inaccurate GPS), historically grown territories with overlap and “gaps”, volatile traffic and time windows, and static plans that were hard to adapt mid‑day.
We delivered an end‑to‑end solution: a time‑dependent, graph‑based routing engine (with multi‑modal options), territory redesign via spatial clustering, and a daily VRP with time windows powered by CP‑SAT and a time‑expanded DAG scheduler. We implemented a rigorous data remediation pipeline (address standardisation, geocoding, deduplication), two‑way CRM integration, and two integration modes (server‑to‑server API for central planning, an embedded SDK for on‑device re‑routing and offline use). We ran an A/B pilot and scaled progressively with dashboards and monitoring to manage by metrics.
Result:
- Approximately 30% less travel time and ~20% more visits per day,
- 15–20% reduction in transport costs,
- Between‑zone travel reduced by around 50% following the territory redesign,
- Improved on‑time performance via real‑time re‑routing and live CRM synchronisation.
From static routes to a dynamic, time‑dependent graph
We modelled the operating area as a weighted, time‑dependent graph with roads, public transport and walking links as edges. Shortest paths are computed with Dijkstra (for pre‑computed matrices) and A* (for fast, on‑the‑fly re‑routing), incorporating live traffic and incidents from mapping APIs. The engine supports multi‑modal choices (e.g., park‑and‑ride + train) where it truly reduces total journey time, not just distance.
Territory design using spatial clustering (OPTICS)
To eliminate spaghetti‑like routes and cross‑territory zig‑zags, we clustered partner locations into contiguous, compact zones using a density‑based method suited to uneven spatial distributions. Business rules then refined clusters (key accounts, regulatory or regional constraints). In practice this halved between‑zone travel and created coherent daily catchments, lifting visit density and agent familiarity with each zone.
Data quality, geocoding and CRM integration
Prior to optimisation, we standardised addresses, repaired or re‑derived missing GPS via geocoding, and resolved inconsistencies between sources. A two‑way integration keeps CRM entities (accounts, visits, constraints) and the routing store in sync, with automated geocoding for new or changed addresses. This remediation was critical: even the best optimisation underperforms with noisy inputs.
Daily VRP with time windows, backed by CP‑SAT and a DAG scheduler
Each planning cycle builds a feasible, near‑optimal sequence of visits per agent under time windows, service durations, working‑hour limits and priorities. Heuristics generate a good initial plan, CP‑SAT and a time‑expanded DAG refine sequencing and enforce constraints efficiently. The approach balances objectives: minimise dead kilometres, meet frequency rules and maximise coverage of high‑value segments.
Real‑time adaptability, API/SDK delivery and measurable impact
Plans adapt during the day: cancellations, urgent inserts or traffic disruptions trigger rapid re‑routing (A*) and, where useful, task swaps across nearby agents. The server API powers centralised overnight planning, the embedded SDK provides on‑device, offline‑tolerant adjustments with in‑app navigation. Territory dashboards expose visit adherence, plan deviations, causes of cancellation, mileage and bottlenecks for continuous improvement.
Contact us
We are equipped with an Agile mindset — flexible, transparent, and results-driven at every stage.
Viacheslav Kostin, CEO
20+ years of experience in managerial positions in IT and banking.
Viacheslav Kostin, CEO
Previous roles: CEO in IT, Director of Strategy and Marketing in Banking, Curator of Holding Banks, Head of Products and Project Office.
Education: MBA for Executives at IMD (Switzerland), Leading Digital Business Transformation (IMD). Provides consulting in strategy and digital transformation.
Vasil Pahomov, CTO
20+ years of experience as a developer, analyst, and solutions architect.
Vasil Pahomov, CTO
Designs resilient, high-load systems with multiple integrations for banks and financial institutions. Expertise in distributed storage and microservices architecture.