Attendees & Representation

TAC Members and Project representatives should mark their attendance below 

Agenda

• Check Action Items & Topic Requests (Backlog)

• Coding Agents and Development

  • Other examples of Agents and Policy: Linux Kernel

  • Legal

  • LF Recommendations - Adapt project documentation and websites to AEO (Answer Engine Optimization) -

• OpenAN Candidate Project

  • Please review and we will vote on the candidate induction on May 13, 2026

• Any Other Topics

LF Staff Updates

@Casey Cain Briefly noted that LF staff are currently attending an all-hands meeting in Utah, which may cause some delayed responses to community inquiries.

AI Agent Ideas and Observability

@Fatih Nar introduced the concept of building an end-to-end reference architecture for applied observability in telecom networks, specifically pipelining observability data into AI machinery. The project aims to identify gaps, create a sandbox, and publish findings and a demo within six months, partnering with operators like Verizon.

• Question from @Olaf Renner (Nokia): How will this collaborate with existing LFN projects, and will it use any of them?

  • Consensus: The initial six-month phase will be tightly scoped to avoid external dependencies, but subsequent iterations next year aim to leverage LFN Super Blueprints and Nephio.

• Question from @Muddasar Ahmed: Can we define a clear roadmap or MVP plan so that it results in a deployable solution rather than just another whitepaper?

  • Consensus: The primary outcome will be a tangible demo using a sandbox with synthetic/real operator data, accompanied by a whitepaper detailing the build process and identified gaps.

• Action Items:

  @Fatih Nar to post the observability presentation to the TAC minutes

  @Fatih Nar to provide more detailed breakdowns and roadmaps for the project phases.

AI Task Force White Paper Principles

@Ranny Haiby introduced a refresher on the AI Task Force's white paper principles, particularly focusing on Principle 5 (agents needing high-quality observability data) and Principle 1 (preparing projects for AI agent contributions). Ranny highlighted that projects are already receiving agent-generated contributions and that shared resources are available on Answer Engine Optimization (AEO) and the Linux kernel's AI contribution policy.

• Question from @Olaf Renner (Nokia): Should projects start exposing agent-native operational interfaces through MCP servers?

  • Consensus: Projects should figure out how to expose functionality to agents, but should avoid locking into a specific technology like MCP, as the industry is shifting rapidly and wrapping complex interfaces (like ORAN) in MCP has caused issues with filling up LLM contexts.

• Question from @Olaf Renner (Nokia): Are any projects, like http://FD.io , already working on this in their roadmaps?

  • Consensus (via @Dave Wallace): Not currently; http://FD.io is still focusing on foundational issues like code attributions and TCO.

• Question from @Dave Wallace: Is the LF formally recommending that projects adopt the Linux kernel AI contribution policy?

  • Consensus (via @Casey Cain): The LF as a whole is not setting a detailed, overarching policy. Instead, the recommendation is for the TAC and individual projects to establish their own policies, using the Linux kernel policy as a strong example.

Demonstrations and Open Source Reference Platforms

@Douglas Knisely introduced the concept of establishing open source reference platforms to act as a baseline for testing AI algorithms. He emphasized the need to focus on tractable, simple use cases (such as energy savings) rather than struggle with overly complex, high-level theoretical issues.

• Action Items:

  Schedule Tracy to present specific use cases (such as the energy saving use case) at the next TAC meeting.

OpenAN Candidate Project

@Casey Cain Introduced a new candidate project proposal called OpenAN, submitted by Huawei. Since Huawei representatives were unable to attend the meeting, the review will be handled offline initially.

• Action Items:

Wiki Migration and Support Requests

@Casey Cain introduced updates regarding LFN's web infrastructure and support protocols. He recently completed a major rewrite and migration of several project wikis (including L3AF, LFN Super Blueprints, Paraglider, StratoWeave, and more) to Confluence.

• Action Items:

  • Project members to notify Casey if they notice any issues resulting from the wiki migrations.

  • Community members must email support@lfnetworking.org as their first step when requesting staff assistance, before pinging staff on Zulip or emailing them directly, to ensure proper ticket tracking.