Pathum Dilshan, RF and Network Planning Engineer

Pathum Dilshan

RF & Network Planning Engineer at Mobitel · Telecom Data & AI · MSc Data Science & AI (University of Moratuwa)

About Pathum Dilshan

Pathum Dilshan is a telecommunications engineer from Colombo, Sri Lanka. He works as an Engineer – Network Planning & Optimization at Mobitel, where he plans and optimizes a live GSM, UMTS, LTE and 5G network and builds data tools that make that work faster.

He holds a BSc (Hons) in Electronic & Telecommunication Engineering from CINEC Campus and is reading for an MSc in Data Science & Artificial Intelligence at the University of Moratuwa. His research focuses on drift-aware anomaly detection and root-cause ranking for mobile RAN KPI data.

He is open to RF/RAN optimization, telecom data & AI, network analytics and data engineering roles, on-site in the UAE, Saudi Arabia, Qatar or Malaysia, or remote.

Experience

Engineer – Network Planning & Optimization, Mobitel (Pvt) Ltd

Sep 2024 – Present · Colombo, Sri Lanka

  • Supported a 500+ site Huawei-to-ZTE vendor swap in Gampaha with minimum service downtime
  • Reduced Call Drop Rate by 15% and increased downlink throughput by 20% across 250+ sites
  • Improved handover success by 12% through neighbour-list optimization and PCI planning
  • Built Python and Apache Airflow pipelines processing 10M+ KPI records a day, cutting manual reporting by about 40%
  • Developed ML-based KPI anomaly detection, improving detection accuracy by 18% over static thresholds

Technical Executive (Huawei), Prime Engineering Lanka

Feb 2024 – Sep 2024 · Deployed to Hutchison Telecommunications Lanka

  • Network analysis and troubleshooting within SLAs on Huawei U2020, ZTE RAN EMS and Nemo
  • Root cause analysis across CRM, EMS, CEM, OSS and HSS data, improving handover success and throughput by 12%

Projects & Research

Drift-Aware Anomaly Detection and Root Cause Ranking for Evolving Mobile RAN KPI Data

MSc research · University of Moratuwa · Supervisor: Prof. Indika Perera

A drift-aware detector that adapts to normal change in network KPIs without learning genuine faults as normal, plus a method that ranks the KPIs most likely to explain each anomaly.

Diabetic Retinopathy Grading with Uncertainty Estimation and Prediction Rejection

MSc project · EfficientNetB0, MC Dropout, Deep Ensembles

Macro AUC above 0.90, calibration error cut by 51%, and 6.3 points higher accuracy on accepted predictions by referring the 23% least-certain cases to a human expert.

Supermarket Sales Dashboard – Looker Studio

Business Intelligence · 817,741 transactions

Interactive dashboard showing sales, margin and customer patterns by district, weekday and age group.

NYC Citi Bike Trip Analysis – Data Visualization

Group project · Python, pandas, matplotlib, seaborn

Visual analysis of 735,502 bike-share trips answering eight questions on usage, stations, riders and trends.

Writing

I Built a Sales Dashboard. Here's What I'd Change.

An honest review of my own Looker Studio dashboard, and seven rules for dashboards people actually use

How I Automated 10M+ Network KPI Records a Day with Python and Apache Airflow

The pattern behind a telecom reporting pipeline, and the mistakes that taught me the most

Swapping 500+ Live Cell Sites Without Customers Noticing: Lessons from a Vendor Swap

What I learned supporting a Huawei-to-ZTE swap on a live mobile network

What 817,741 Supermarket Transactions Taught Me About Reading Data

Five findings from a Looker Studio dashboard, and the one that wasn't a finding at all

From Drive Tests to Data Pipelines: Why I Went from RF Engineer to Data & AI

What carried over, what didn't, and what I'd tell any engineer thinking about the same move

Your Network Alarms Are Lying to You

Why static KPI thresholds fail in mobile networks, and the drift trap that catches even ML models

I Built a Medical AI Model That Knows When to Say “I’m Not Sure”

Medium · Uncertainty estimation and prediction rejection for safer clinical AI

I Built Three Reinforcement Learning Agents. The “Improved” Ones Lost.

Medium · Variance, baselines and seeds in policy-gradient methods

Skills

RF & Radio Network OptimizationLTE / 5G RANHuawei U2020ZTE NetNumen U31PythonSQLApache AirflowMachine LearningDeep LearningAnomaly DetectionTime Series AnalysisPower BILooker StudioAWS

Education

MSc in Data Science & Artificial Intelligence (reading)

University of Moratuwa, Department of Computer Science & Engineering · 2026 – 2027

BSc (Hons) Engineering in Electronic & Telecommunication Engineering

CINEC Campus, Malabe

FAQ

Who is Pathum Dilshan, the telecom engineer?

Pathum Dilshan is an RF and network planning engineer at Mobitel (Pvt) Ltd in Colombo, Sri Lanka. He plans and optimises a live 2G/3G/4G/5G mobile network and builds data engineering and machine learning tools for network KPIs.

Is Pathum Dilshan the engineer the same person as Pathum Dilshan the cricketer?

No. This Pathum Dilshan is a telecom and data/AI engineer at Mobitel. He is a different person from Pathum Dilshan (born 1991), the Sri Lankan first-class cricketer.

What has Pathum Dilshan worked on?

He supported a Huawei-to-ZTE vendor swap of 500+ live cell sites, led optimisation across 250+ sites that cut call drop rate by 15% and raised downlink throughput by 20%, built Python and Apache Airflow pipelines processing 10M+ KPI records a day, and developed an ML-based KPI anomaly detection approach that improved detection accuracy by about 18% over static thresholds.

Where did Pathum Dilshan study?

He holds a BSc (Hons) in Electronic & Telecommunication Engineering from CINEC Campus and is reading for an MSc in Data Science & Artificial Intelligence at the University of Moratuwa, where his research focuses on drift-aware anomaly detection and root-cause ranking for mobile RAN KPI data.

How can I contact Pathum Dilshan?

Email pmpdilshan@gmail.com or connect on LinkedIn at linkedin.com/in/pathum-dilshan.

Contact

Email pmpdilshan@gmail.com or connect on LinkedIn.