MGMilind Gunjan / gunjanlabs

About

Telecom engineer, building trustworthy AI for networks.

I started out tuning LTE radios. Fourteen years later I build the controls that make AI safe to put inside carrier networks: inference that holds up under load, agents that stay inside policy, and evidence before any action touches the network.

The short version

I began in radio planning and optimization at Ericsson, then led indoor LTE small-cell deployments. At Sprint and T-Mobile I spent seven years on the shift from boxes to software: a multi-vendor NFV testbed behind one of the first NFV-based VoLTE launches, a 5G standalone core lab, the OMEC open-source packet gateway with Intel, and Kubernetes for network functions.

At Netskope I turned that into security: clientless Zero Trust for cellular IoT devices that can't run an agent, rooted in the SIM. That work produced three of my four granted patents.

Today I'm a Technical Account Manager at AWS, working with a Tier-1 U.S. operator and in AWS's telecom group on 6G and AI-native networks. Outside the day job I build in the open: the AI-Native ODA Canvas contract for TM Forum, agentic Kubernetes guardrails for ISC2 Security Congress, cross-layer assurance for IEEE CQR, and the inference lab on this site.

This site and everything on it are my own. Views are not my employer's, and nothing here describes a customer.

What I believe

Telco AI is the road to autonomous networks

Not a shortcut around operations. Agents assist; humans and policy stay in charge until the evidence says otherwise.

Evidence before action

No agent touches a network without a gate it has to pass: tests, telemetry, a reviewer, a rollback.

Build the lab first

Every claim here maps to something you can run, read or check: a repo, a simulation, a talk, a patent.

Label what it is

Illustrative is labeled illustrative. Pending is labeled pending. A hackathon is a hackathon.

Timeline

  1. 2026Four stages, a fourth patent, the inference lab

    ISC2 Security Congress, IEEE CQR, TM Forum DTW Ignite and Innovate Americas. US 12,757,023 granted. LLM inference series published here.

  2. 2025Zero Trust for cellular, in public

    CSA Zero Trust for Cellular Networks focus group; Forbes Technology Council article on Salt Typhoon; two Netskope patents granted.

  3. 2024 – nowAWS · Technical Account Manager, telecom

    Technical lead on the AWS side for a Tier-1 U.S. operator; AWS telecom group on 6G and AI-native networks.

  4. 2022 – 2024Netskope · 5G Edge Solutions Architect

    Clientless, SIM-rooted Zero Trust for cellular IoT and OT; multi-vendor 5G security testbed.

  5. 20195G Americas, 5G and the Cloud

    Co-led the industry white paper on cloud-native 5G core.

  6. 2015 – 2022Sprint / T-Mobile · Senior Technology Development Strategist

    NFV testbed behind one of the first NFV-based VoLTE launches; 5G SA core lab; OMEC with Intel; Kubernetes for a 5G open innovation lab; SCP patent.

  7. 2013 – 2014QuadGen Wireless · Senior Network Engineer

    Led indoor LTE small-cell deployments end to end.

  8. 2010 – 2013Ericsson India · Network Engineer

    LTE radio planning and optimization for a U.S. Tier-1 network.

Patents

PatentTitleGranted
US 12,757,023Cellular IoT security using dynamic policy-driven mechanisms for threat detection and isolationOct 2026
US 12,245,036Global secure SIM clientless SASE architecture for cellular devicesMar 2025
US 12,219,360Cellular IoT security using dynamic policy-driven mechanisms for threat mitigationFeb 2025
US 12,155,534Service communication proxy (SCP) that serves network topology dataNov 2024

Pending

PublicationTitleStatus
US 2026/0019811 A1Clientless SASE architecture with adaptive proxy policies enforcement based on certificate installationPublished Jan 2026, pending

All co-invented: the SCP patent at T-Mobile, the others at Netskope. Citations are tracked on Google Scholar.

Standards and community

  • Vice Chair, IEEE Systems Council Technical Committee on System Security
  • Activity Chair, IEEE-SA 5G/6G Security
  • Member, CSA Zero Trust Working Group, and its Zero Trust for Cellular Networks focus group
  • Project Participant, TM Forum AI-Native ODA Canvas Innovation Hub
  • Peer reviewer, Computers & Security (Elsevier), 2026

Certifications

CISSP · NVIDIA-Certified Associate: AI Infrastructure and Operations · AWS Certified Machine Learning Engineer – Associate · AWS Certified Solutions Architect – Associate · AWS Certified AI Practitioner

Education

M.S., Telecommunications, University of Maryland, College Park · B.E., Electronics & Communication Engineering, Manipal Institute of Technology

Elsewhere

GitHub · LinkedIn · Google Scholar · Medium