Mohit Sharma
Menu

Projects / Enterprise

Enterprise & Professional Work

This section brings together work delivered in enterprise environments where constraints are real, systems must operate at scale, and outcomes carry measurable impact.

The focus here is not only on models, but on how those models fit into decision systems, how they interact with operational processes, and how they perform under real-world conditions.

Case studies

Telecom · Forecasting · ML Systems

Inventory Optimisation & Forecasting System

Designing a forecasting system for inventory planning under real-world constraints, balancing accuracy, scalability, and operational usability.

Impact

70 MSEK realised savings · 350 MSEK projected impact

ForecastingSupply ChainDecision Systems
View Case Study →

Banking · Fraud Detection · Decision Systems

Credit Card Fraud Detection

Designing a fraud detection system for rare-event environments, combining supervised models, anomaly detection, and decision-layer controls.

Impact

Improved detection of rare fraud events while maintaining operational control over false positives

BankingFraud DetectionRisk Systems
View Case Study →

Banking · Growth Analytics · Customer Targeting

Propensity & Cross-Sell Modelling

Developing targeted modelling systems to identify high-probability customer opportunities within a regulated financial environment.

Impact

750K+ targeted opportunities identified

Customer AnalyticsGrowthML Models
View Case Study →

Technology · NLP · Large-Scale Systems

Semantic Understanding & Topic Modelling

Building semantic understanding systems using topic modelling, clustering, and similarity techniques across large-scale text data.

Impact

Applied across large-scale language and search datasets

NLPTopic ModellingSearch Systems
View Case Study →

A note on this work

Enterprise work is shaped by constraints that do not exist in controlled environments. Data is imperfect, systems are complex, and decisions have real consequences.

The value of this work lies not only in building models, but in designing systems that behave reliably under those conditions. That includes understanding trade-offs, aligning with operational realities, and ensuring that outputs translate into meaningful decisions.