About Expertise Work Let's Talk →
Telecom Core Network Engineer

Ankit
Jha

Telecom Core & IoT Engineer   |   DIAMETER  ·  AAA  ·  Kafka  ·  Distributed Systems

I build the infrastructure that keeps telecom networks running — authentication systems, policy control, compliance tooling, and distributed data pipelines that operators depend on at scale.

5G+
Protocol Stacks
IPLMS
DoT / TRAI Ready
Gx/Gy
Core Interfaces
ACCESS EPC / 5G CORE SPECIALISATION Uu N2 N3 Diameter Gx / Gy Gy N6 UE gNB / RAN AMF SMF UPF AAA PCRF HSS / UDM IPLMS Internet Data path DIAMETER / AAA Specialisation 3GPP EPC / 5GC REFERENCE
About

Engineering Networks at Scale

I'm a Telecom Core Network Engineer specialising in the critical, invisible layer that every call, session, and byte of data passes through — authentication, policy enforcement, and compliance reporting. If a network can't correctly identify, authorise, or log its subscribers, everything downstream fails.

My most recent build was an IPLMS platform from the ground up, alongside DIAMETER AAA implementation for a telecom startup — handling real production traffic with regulatory compliance requirements architected in from day one, not bolted on after.

I also build distributed data platforms and IoT systems — streaming pipelines with Kafka, large-scale processing with PySpark and Databricks, and microservice architectures that handle the volume and velocity telecom and IoT environments actually produce. If your network generates data faster than you can process it, that's a problem I've solved before.

CFA Programme — Currently Pursuing

Most engineers understand how infrastructure works. The CFA programme is teaching me why capital flows where it does — which investments operators will actually make, which SLA commitments are financially sustainable, and where the next hundred billion in telecom and IoT infrastructure will be spent. Network engineering and capital allocation intersect more than either profession admits. Understanding both sides of that table is a deliberate choice, not a distraction.

Core Expertise

What I Build

01
DIAMETER / AAA Protocol

Full implementation of the DIAMETER protocol stack covering authentication, authorisation, and accounting across Gx, Gy, and Rx reference interfaces — built for operators who can't afford sessions to drop.

GxGyRxRADIUS
02
IP Log Management (IPLMS)

End-to-end IPLMS platforms for DoT and TRAI regulatory compliance. Built from scratch, handling production-scale IP log capture, storage, and reporting — ready for inspection, every time.

DoTTRAICompliance
03
5G AAA & Policy Control

Core network functions for 5G deployments — AUSF, UDM, and PCRF implementation with policy control that scales with subscriber growth without manual intervention.

AUSFUDMPCRF5G SA
04
MVNO Core Architecture

Core network design and implementation for MVNOs — from subscriber provisioning to roaming interconnect. Helping virtual operators launch quickly without compromising on network integrity.

MVNOHLR / HSSRoaming
05
ISP Network Infrastructure

Core and access layer infrastructure for ISPs — AAA for broadband subscribers, billing integration, and network policy enforcement at the edge.

Broadband AAACGNATBilling
06
IoT & M2M Platforms

End-to-end IoT platform design — device connectivity, AAA for machine subscribers, real-time telemetry ingestion, and session management at M2M scale. Built for operators running thousands of connected devices under a single network policy.

IoTM2MDevice AAATelemetry
07
Distributed Data & Streaming

Real-time data pipelines built on Kafka for high-throughput telecom event streams — CDRs, session events, policy triggers — processed and enriched at scale using PySpark and Databricks. Microservice architectures that decouple ingestion from analytics without sacrificing latency.

KafkaPySparkDatabricksStreaming
08
AI Analytics · Telecom & IoT

Machine learning pipelines that turn raw network telemetry and IoT data into operational intelligence — anomaly detection, traffic forecasting, and subscriber behaviour modelling built on distributed compute.

ML PipelinesAnomaly DetectionPySparkDatabricks
Services

How I Can Help

🏗️
Core Network Design & Build

Architecture and implementation of telecom core network elements from the ground up — whether you're launching an MVNO, upgrading an ISP, or deploying private 5G.

  • DIAMETER / AAA stack implementation
  • HSS / HLR provisioning
  • PCRF & policy control setup
  • End-to-end core network testing
📜
Regulatory Compliance (IPLMS)

Build or audit your IP Log Management System to meet DoT and TRAI requirements. Production-grade platforms designed for operator-scale volumes — not tick-box solutions.

  • IPLMS platform build / overhaul
  • DoT & TRAI compliance auditing
  • Log retention & reporting pipelines
  • Readiness assessment
📊
AI-Powered Network Analytics

Turn raw network telemetry and IoT data into actionable intelligence — traffic insights, anomaly detection, and subscriber behaviour modelling.

  • Network telemetry pipelines
  • Anomaly & fraud detection
  • Subscriber behaviour modelling
  • Custom dashboards & reporting
🔍
Technical Advisory & Review

An independent review of your core network architecture, AAA configuration, or compliance posture — with clear, actionable recommendations from someone who has built it in production.

  • Architecture review & gap analysis
  • AAA configuration audit
  • Vendor selection support
  • 5G migration planning
Featured Work

Selected Projects

N12 N7 N8 AMF AUSF PCRF UDM 5G STANDALONE CORE
5G Core Private 5G
Telecom

5G AAA & Policy Control

Core network architecture for a private 5G deployment — AUSF, UDM, and PCRF implementation with policy control designed to scale without manual intervention.

AUSFUDMPCRF5G SA
TELEMETRY ML Pipeline Anomaly · Forecast INSIGHTS AI-POWERED NETWORK ANALYTICS
AI Analytics Telecom / IoT
Telecom / IoT

AI-Powered Network Analytics

Network operators were accumulating years of telemetry from telecom and IoT sources, but data was processed in batch and surfaced through static dashboards that told you what happened — not what was about to. I built a streaming pipeline on Kafka for real-time telemetry ingestion, PySpark Structured Streaming for enrichment and feature engineering, and Databricks as the compute and storage layer — with anomaly detection running continuously on live KPI streams and a traffic forecasting model producing 24-hour load predictions per network segment. Operators can now see degradation patterns hours before user impact, and capacity decisions are driven by forward-looking forecasts rather than reactive incident reports.

KafkaPySparkDatabricksIoT
"sub-5ms for500 robots" AI IntentLLM + RAG3GPP Rel-17/18 5G SliceS-NSSAI · 5QI · AMBR Enterprise Provisioned in <60s 5G SLICING MONETIZATION PLATFORM
5G · AIPlatform
Telecom

5G Network Slicing Monetization Platform

Operators have invested $500B+ in 5G but ARPU remains flat because connectivity is still sold as a commodity. This platform lets enterprises describe connectivity needs in plain English — an LLM+RAG engine (trained over 3GPP Rel-17/18 specs) translates intent directly into S-NSSAI, 5QI, AMBR, and GFBR parameters, provisioning a dedicated slice in under 60 seconds. SLA Assurance AI detects breach risk sub-minute; a Contextual Bandit RL agent optimises pricing per slice tier; CAMARA-compliant APIs expose slice control to third-party developers.

5G SANetwork SlicingLLM + RAGNEFCAMARA
BMS HVAC NMS AI CoreLSTM · GNNLLM+RAG "HVAC-3B bearingwear — fix in 90s" Auto Work Order IoT OPERATIONS INTELLIGENCE
IoT · AIPlatform
IoT

IoT Operations Intelligence Platform

Facility managers and NOC teams work from siloed systems that generate raw alarms without context — manual fault investigations take 40–60 minutes per incident. This platform unifies BMS sensors (HVAC, energy, access, elevators) and NMS devices (routers, BTS/RAN, servers, UPS) under a single Chat-to-Fix interface: operators describe a fault in plain English, the LLM+RAG engine retrieves device history and vendor runbooks, and returns a specific resolution in under 90 seconds. An LSTM model predicts hardware failures 24–72 hours ahead with 87% precision and auto-generates prioritised work orders; a GNN traces alarm cascades to the single root cause in under 8 seconds.

MQTTKafkaLLM + RAGLSTMGNNBACnet
RAN KPIs IPLMS Weather NMS Alarms 4-Model MLLSTM·GNN·RFIsolation Forest 91 HIGH RISK Cell A · 72hr window AI RANGE DETECTION & OUTAGE PREDICTION
AI · TelecomPlatform
Telecom

AI Range Detection & Outage Prediction

Telecom operators are entirely reactive — they learn about coverage failures when users complain, after the damage is done. This platform inverts that model: a 4-model ML ensemble (LSTM/TFT for time-series KPI forecasting, GNN for topology cascade propagation, Spatial RF Model for coverage hole prediction, Isolation Forest for real-time anomaly detection) fuses RAN KPIs, IPLMS session flows, weather/terrain data, UE probes, and NMS alarm history to produce a 0–100 outage risk score per cell site every hour. SHAP explainability surfaces the top 3 contributing risk factors; an LLM narrates the risk in plain English and recommends a proactive intervention — before a single user drops a call.

LSTM/TFTGNNXGBoostSHAPKafkaPySpark
5G KPIs CDRs SLA Data Financials Kafka DeltaLake Revenue Impact Churn Prediction Exec NL Chat SLA Financials AURA — AI UNIFIED REAL-TIME ANALYTICS
BI · AIPlatform
Telecom / BI

AURA — AI Unified Real-time Analytics for 5G

Telecom operators hold some of the richest data estates in the enterprise world — 5G KPI telemetry, CDRs covering hundreds of millions of subscribers, SLA logs, financial datasets — yet the vast majority never reaches the executives who need to make capital allocation decisions. AURA consolidates all network and business data into a Databricks Delta Lake lakehouse (Bronze/Silver/Gold), processed in near real-time by Apache Spark, with five AI models running continuously: XGBoost+LSTM churn prediction, TFT ARPU forecasting, Causal ML revenue attribution, Monte Carlo capacity ROI simulation, and a RAG-grounded executive LLM that answers any business question with auto-generated charts in under 10 seconds.

DatabricksKafkaPySparkDelta LakeCausal MLMLflow
Who I Work With

Building for the Operators Who Keep the World Connected

If you're building an MVNO, running an ISP, or deploying private 5G infrastructure, you need a core network engineer who has done it before — in production, under compliance scrutiny, with real traffic.

I'm open to freelance engagements, project-based consulting, and longer-term technical partnerships.

Mobile Virtual Network Operators (MVNOs)
Internet Service Providers (ISPs)
Private 5G Deployments
Telecom Startups & New Entrants
Regulatory & Compliance Projects
IoT & M2M Platforms
Open to Freelance

Let's Build Something
That Scales.

Whether you're building telecom core infrastructure, designing an IoT data platform, or need expertise across DIAMETER, 5G, Kafka, or distributed systems — I'd like to hear what you're working on.

ankitjha8891@gmail.com