Behzad Kianbakht

2020

MTN Irancell

Software and Automation Senior Specialist

Present

Jul 2020 — Present • 5+ years

Built and operated a production AIOps platform at telco scale — combining platform engineering, infrastructure automation, CI/CD pipelines, IaC frameworks, and ML-driven incident prediction across 1,500+ critical business services on a 5,000+ host production estate..

Designed and shipped a Kubernetes-native internal observability platform (RKE2/Rancher, Helm) consolidating Prometheus, Grafana, ELK Stack, and Zabbix into a single pane of glass — reducing platform provisioning time by 80% and enabling engineering teams to onboard new services without ops intervention.
Designed and built a full-stack AIOps platform (Service-Oriented AIOps) — a 9-layer production system collecting infrastructure and service telemetry from Zabbix (Infra + Service) and Jira ITSM via Python 3 collector modules, persisting raw data into a PostgreSQL schema and serving a FastAPI + Bootstrap web dashboard with per-service incident probability, risk scoring, and health monitoring across Payment, SMSC, IN, and CRM business services.
- ↳ML pipeline: Engineered service-based features via RapidMiner (correlation analysis, information gain, PCA, time-series aggregation) into a dedicated feature store — trained and evaluated Random Forest, XGBoost, SVM, and Neural Network classifiers with full model registry and versioning, delivering risk levels (Low / Medium / High / Critical) with real-time alerting via Telegram, Slack, and Webhook.
- ↳Prediction engine: Delivered per-service incident probability scores, anomaly detection, and health scores — enabling proactive intervention across 85% of high-business-impact failure scenarios before user-facing degradation.
Architected and developed a modular data source integration framework in Python — designed pluggable connectors for heterogeneous sources including service-based APIs (REST, SOAP), database adapters (MySQL, MongoDB, Oracle), and protocol-level collectors (SNMP) — each normalising raw data into structured monitoring metrics across 230+ services. Automated versioning and deployment via GitLab CI/CD with GitLab Runner, supporting both Docker-based and Bash-based pipeline executors.
Defined and enforced SLO/error budget contracts across 1,500+ services — built tooling to track burn rates in Grafana, trigger automated error budget alerts, and feed ML model retraining cycles, reducing major incidents by 40% and moving the team from reactive firefighting to data-driven reliability management.
Engineered an intelligent incident automation system integrating the monitoring stack with Jira API — designed an event-driven pipeline (alert → classification → priority routing → ticket creation → SLA tracking → auto-closure) that eliminated manual toil and significantly reduced MTTR.
Delivered an Infrastructure as Code platform using Terraform and Ansible — designed reusable module libraries, enforced GitOps workflows via GitLab CI/CD, and achieved zero configuration drift across hybrid cloud and on-premises environments.

▲ 80% provisioning savings ▲ 40% fewer incidents ▲ 85% AI risk coverage ▲ 5,000+ hosts ▲ 1,500+ services

KubernetesRKE2Rancher HelmTerraformAnsible PrometheusGrafanaELK GitLab CI/CDPython 3.12PostgreSQL RapidMinerXGBoostFastAPI AIOpsZabbixJira ITSM

2019

Amin Optical Transmission Network

Software Engineer

Networking & ML

Jan 2019 — Jul 2020 • 18 months

Shifted from hardware-level networking to software-defined networking, building ML-driven automation for optical network management.

Built Python-based automated topology discovery for 250+ network nodes and links, exposed via FastAPI for NMS/OSS integration
Developed ML-driven routing & wavelength assignment engine (ILP optimisation) — 70% improvement in resource utilisation
Implemented SDN controller with Northbound RESTCONF and Southbound NETCONF/OpenFlow interfaces — eliminated CLI-based configuration

▲ 70% resource efficiency ▲ 250+ nodes automated

PythonFastAPISDN RESTCONFNETCONFOpenFlow scikit-learnILP

2017

FiberHome Telecommunication Technologies

Software Developer

Network Software

Jun 2017 — Mar 2019 • 21 months

Deep systems programming for optical network switching — routing algorithm design and protocol module optimisation.

Developed concurrent fault-tolerant routing algorithm in embedded C for ASON optical platform — sustained 99.9% uptime
Refactored OSPF and CSPF routing modules — achieved 40% throughput improvement and 20% latency reduction

▲ 99.9% uptime ▲ 40% throughput ▼ 20% latency

Embedded COSPFCSPF ASONNetwork Algorithms

2015

Electronics Company

Embedded Systems Designer

Foundation

Aug 2015 — Feb 2017 • 18 months

Built the engineering foundation — hardware design, signal processing, and low-level firmware for aviation-grade systems.

Designed aviation-grade Altitude Encoder (MSP-430) with barometric pressure processing for Mode C transponder compliance
Developed TCP/UDP protocol stacks on FPGA (Wiznet 5500) and Air Data Computer in VHDL/MicroBlaze C++
Delivered 10-layer PCB in Altium Designer, validated with HyperLynx signal integrity analysis

MSP-430FPGAVHDL C++AltiumPCB Design

2008

Academic Foundation

B.Sc. Electronic Engineering → M.Sc. Telecommunication Engineering

Education

2008 — 2020 • Azad University

From undergraduate electronics to postgraduate telecom research — culminating in an IEEE-published paper on Massive MIMO resource allocation.

M.Sc. 16.77/20B.Sc. 15.48/20 IEEE Publication ↗Cell-Free Massive MIMOICEE 2020

From Embedded Systems to Software Engineering

Core competencies

Academic background

📄 Download My Resume

Let's talk

Available for new opportunities