CNCF celebrates profitable mentees from LFX Mentorship 2025 Time period 3

The Cloud Native Computing Basis is thrilled to congratulate the 2025 Time period 3 (September – November) CNCF LFX Mentorship Program mentees who’ve efficiently accomplished this system! This time period noticed a incredible cohort of mentees working throughout 27 totally different Graduated, Incubating, and Sandbox tasks, together with Cilium, KubeArmor, KubeEdge, Kubernetes, OpenCost, and lots of extra.

Our mentors and mentees have labored tirelessly over the previous three months to enhance code, documentation, and group engagement. The LFX Mentorship program continues to be a significant pipeline for brand new contributors to the cloud native ecosystem, fostering the subsequent technology of open supply leaders.

Under, you’ll discover highlights from a few of our graduating mentees, adopted by a full listing of all of the profitable tasks and individuals from this time period.

The subsequent time period of CNCF LFX Mentorship is accepting functions now! For those who’re excited by taking part, take a look at our listing of tasks to use to.

Functions are open til February 10, 2026.

Mentee Profile Highlights

KubeEdge

Throughout this LFX Mentorship with KubeEdge, I systematically restored the lifelong studying semantic segmentation benchmark in Ianvs—a distributed synergy AI benchmarking toolkit for edge computing. The challenge concerned resolving 17 important points throughout a number of framework layers, together with compatibility fixes for PyTorch 2.0+, CPU-only operation enablement, multi-round data persistence, and path decision for cross-environment deployment.

Mentee: Abhishek Kumar (learn Abhishek’s private weblog publish in regards to the mentorship expertise)

Mentors: Zimu Zheng, Shijing Hu

“My LFX Mentorship with CNCF and KubeEdge has been an extremely enriching expertise that remodeled the way in which I perceive open-source engineering. Over the course of this system, I had the chance to work on a difficult real-world drawback, collaborate intently with skilled mentors, and contribute meaningfully to a production-grade challenge. The steering, evaluations, and steady help from the group helped me develop not simply technically—by way of debugging, restoring advanced pipelines, and understanding large-scale system structure—but additionally professionally, by studying open-source greatest practices, communication, and collaborative improvement. This mentorship has been probably the most impactful studying journeys of my profession, and I’m really grateful for the chance to contribute to CNCF’s vibrant and supportive ecosystem.”

OpenCost

Throughout my LFX Fall 2025 mentorship with OpenCost, I designed and carried out a Mannequin Context Protocol (MCP) server integration that allows AI brokers to question Kubernetes value allocation knowledge by way of pure language. The implementation contains three main instruments built-in straight into the primary OpenCost binary and enabled by default in v1.118+.

Mentee: Adesh Pal (learn Adesh’s private weblog publish in regards to the mentorship expertise)

Mentors: Alex Meijer and Matt Bolt

“The LFX Mentorship program exceeded my expectations in each method. Working with mentors Alex Meijer and Matt Bolt, I discovered not simply how one can write code, however how to consider manufacturing programs—they challenged me on edge instances, pushed me past ‘simply working’ to production-ready, and taught me that nice open supply is as a lot about maintainability and consumer expertise as performance. The structured but versatile program gave me house to discover whereas offering clear milestones that saved me accountable. What stunned me most was how welcoming the OpenCost group was—attending biweekly conferences, I noticed choices being made collaboratively and felt my contributions genuinely mattered.”

WasmEdge

The challenge primarily concerned including help for stateful /responses endpoint to the llama-nexus challenge which on the time supported /chat/completions endpoint which is stateless. The additional objective of the challenge was so as to add help for a code interpreter. This interpreter was run utilizing a sandboxed setting through Docker-container.

Mentee: Ashish Dalal (learn Ashish’s private weblog publish in regards to the mentorship expertise)

Mentor: Michael Yuan

“The expertise with LFX mentorship was very clean and nurturing. The mentors and maintainers had been very supportive all through the challenge and I might take time to check my very own concepts and approaches to a function. Options given throughout PR overview proved very useful and at occasions insightful. Every time I confronted challenges in the course of the mentorship, I used to be given regular help and assist.”

Kmesh

Partaking with the open-source group and fixing issues, particularly these associated to implementing elements of IPSec, has been an immensely rewarding and fulfilling expertise. All through the challenge, my mentors guided me in driving the progress of the open-source initiative, patiently answering my questions and offering course.

Mentee: Haobin Huang (learn Haobin’s private weblog publish in regards to the mentorship expertise).

Mentors: Zhonghu Xu, Zhencheng Li

“General, I had a really optimistic expertise. The primary problem was that I discovered it fairly troublesome to get began and orient myself throughout the total challenge on the very starting. I really feel extremely lucky to have participated within the LFX Mentorship challenge, because it has considerably enhanced each my data and coding expertise. Earlier than becoming a member of the Kmesh challenge, I had no clear understanding of what a service mesh truly does. By my involvement, I’ve gained a broad overview of the service mesh subject and had the chance to study and get hands-on expertise with cutting-edge developments.”

•picture not out there

Kubernetes

Throughout my LFX Mentorship, I labored on bettering the Kubernetes API reference documentation generator—an important however fragile device used on the finish of each Kubernetes launch cycle. The generator usually failed silently or in undocumented methods, making the method depending on a single professional with deep inner data. My challenge centered on lowering this fragility by refactoring brittle elements of the codebase, including clearer and extra actionable error dealing with, and enhancing the device’s documentation.

Mentee: Lavish Pal (learn Lavish’s private weblog publish in regards to the mentorship expertise)

Mentors: Kat Cosgrove, Rey Lejano, Xander Grzywinski, Nate Waddington

“The educational expertise I had over the past 3 months will show to be useful all through my profession. Going by way of documentation, searching for references, testing issues out and so forth are some expertise that I used to be in a position to enhance throughout this mentorship. The chance to contribute to an actual world challenge as a scholar feels superb. The help from my mentors Nate Waddington, Kat Cosgrove, Rey Lejano made the mentorship expertise productive and fulfilling.”

KubeStellar

My LFX Mentorship with CNCF KubeStellar was a mission to make reliability mission-critical for our multi-cluster Kubernetes platform. I designed and constructed a sturdy, end-to-end Playwright testing framework from scratch, leveraging the Web page Object Mannequin (POM) structure to validate each advanced, distributed system workflow. This effort wasn’t nearly overlaying code; it was about stopping failure, resulting in the prevention of over 25 important, production-bound bugs.

Mentee: Nupur Shivani (learn Nupur’s private weblog publish in regards to the mentorship expertise)

Mentors: Andy Anderson, Onkar Shelke, Shivam Kumar, Rishi Mondal

“My LFX Mentorship with CNCF KubeStellar was a transformative and unparalleled expertise that launched my profession from educational idea to impactful manufacturing engineering. It wasn’t nearly coding; it was a rigorous journey in professionalizing my talent set, demanding hard-earned confidence in architecting advanced options, mastering the CI/CD lifecycle, and stopping regressions in a reside, distributed system. The mentorship construction—combining high-impact tasks just like the E2E testing framework with the surprising alternative to steer and mentor different contributors—supplied progress that far exceeded any classroom.”

Cilium

Throughout my LFX Mentorship with Cilium, I labored on bettering the challenge’s total discoverability and search efficiency by enhancing its search engine marketing, AEO, and AIO. My contributions centered on refining metadata, bettering structured knowledge, reorganizing heading hierarchies, making a devoted FAQ web page, including TL;DR summaries, and guaranteeing machine-readable content material throughout key documentation and web site pages.

Mentee: Peace Sandy (learn Peace’s private weblog publish in regards to the mentorship expertise)

Mentor: Invoice Mulligan

“My LFX Mentorship expertise was extremely rewarding and transformative. This system allowed me to work intently with skilled maintainers, achieve hands-on publicity to actual open-source improvement, and strengthen each my technical and communication expertise. The help from my mentor and the Cilium group made studying fulfilling, collaborative, and deeply empowering. This mentorship not solely expanded my understanding of cloud-native applied sciences but additionally boosted my confidence to contribute meaningfully to open-source tasks.”

Kubernetes (Kube State Metrics)

Throughout my LFX Mentorship, I centered on totally automating the discharge course of for kube-state-metrics (KSM), which beforehand required lots of handbook effort each few months. I constructed a set of GitHub workflows that streamline your complete pipeline end-to-end automating all instructions wanted to create the pre-release PR, producing pre-releases with GoReleaser, creating launch notes, and dealing with the post-release steps as properly.

Mentee: Rishab Kumar Jha (learn Rishab’s private weblog publish in regards to the mentorship expertise)

Mentors: Pranshu Srivastava, Manuel Rüger

“My total LFX Mentorship expertise was genuinely superior. My mentors had been extremely supportive and at all times out there to information me at any time when I obtained caught, which made your complete journey clean and fulfilling. I discovered rather a lot particularly round launch engineering, CI/CD workflows, GitHub Actions, and the form of real-world practices groups comply with when reducing and managing releases. It was the form of hands-on, industry-level publicity I hoped for, and I actually don’t have any complaints, simply lots of gratitude for the way a lot I discovered and the way welcoming the group was.”

KubeEdge

My challenge could be probably the most attention-grabbing and distinctive LFX tasks this yr. In contrast to many contributions that focus primarily on including code, bettering CI/CD, or writing unit checks, this work was a whole research-driven challenge constructed round Edge AI and Deep Studying. The challenge explores Embodied Intelligence, which is undoubtedly probably the most trending and impactful areas of AI as we speak, with functions throughout each main {industry}.

Mentee: Ronak Raj (learn Ronak’s private weblog publish in regards to the mentorship expertise)

Mentors: Zimu Zheng, Shijing Hu

“From a 16-year-old studying Python to a 19-year-old contributing to cutting-edge AI analysis — this journey has been unbelievable. The LFX Mentorship Program gave me way over technical expertise; it gave me confidence, mentorship, worthwhile connections, and a transparent course for my profession. To my mentors, Dr. Zimu Zheng and Dr. Shijing Hu — thanks for believing in me, guiding me, and continually pushing me to ship my greatest work. Your suggestions, persistence, and help made all of the distinction.”

KubeArmor

Throughout my LFX Mentorship with KubeArmor, I labored on enhancing the challenge’s observability by implementing Prometheus metrics for coverage and alert monitoring. The primary objective was to offer customers higher visibility into how their safety insurance policies are performing in manufacturing. I added 4 key metrics that monitor coverage counts, coverage metadata, complete alerts, and rule violations with detailed labels displaying which insurance policies are being triggered and what actions are being taken.

Mentee: Saurav Teli (learn Saurav’s private weblog publish in regards to the mentorship expertise)

Mentors: Rishabh Soni, Aryan Sharma, Ramakant Sharma, Barun Acharya

“My LFX Mentorship expertise with KubeArmor was extremely worthwhile for studying how real-world open supply improvement works. The mentors had been supportive and affected person, particularly after I hit infrastructure challenges throughout native testing that took days to troubleshoot. What I appreciated most was the steering on writing clear, maintainable code – they pushed me to refactor my preliminary implementation of the rule violation metric from a easy 17-line strategy to a extra sturdy 41-line model with correct filtering logic…. The expertise of contributing to a CNCF challenge gave me confidence in tackling advanced codebases and confirmed me that even when issues don’t work completely the primary time, persistence and asking the suitable questions will get you there.”

Podman Container Instruments

My LFX Mentorship challenge with Podman aimed to resolve a selected UDP networking concern the place Conntrack creates a stale file of a failed connection try, blocking the sender on that port till the entry occasions out. The objective was to implement conntrack entry clearing into Netavark, Podman’s community administration device.

Mentee: Shivang Ok Raghuvanshi (learn Shivang’s private weblog publish in regards to the mentorship expertise)

Mentors: Matthew Heon, Paul Holzinger

“General, it was an ideal expertise. I discovered lots of issues from it. Past the technical expertise, I actually appreciated how welcoming and useful the Podman workforce was. My mentors, Paul and Matt, had been extremely useful and responsive. We had weekly conferences, and I actually recognize the time they devoted to me. Truthfully, they’re a number of the greatest mentors anybody might ask for; their help saved my motivation excessive all through this system.”

OpenCost

Throughout the LFX Fall 2025 mentorship with OpenCost, I labored on KubeModel—a next-generation knowledge mannequin designed to deliver extra correct and dependable value monitoring to Kubernetes environments. The core drawback I tackled was that Kubernetes is inherently dynamic: pods get recreated, names get reused, and assets continually shift, which makes conventional name-based monitoring unreliable.

Mentee: Sparsh Raj (learn Sparsh’s private weblog publish in regards to the mentorship expertise)

Mentors: Alex Meijer, Sean Holcomb, Niko Kovacevic

“My LFX Mentorship expertise with OpenCost was genuinely rewarding. What I appreciated most was being concerned within the design course of from the bottom up – slightly than simply implementing predefined duties, I obtained to take part in architectural discussions and assist form how KubeModel would work. My mentors, Alex, Sean, and Niko, had been extremely supportive all through. They gave me the house to discover concepts whereas offering steering after I wanted it, they usually had been at all times open to questions and suggestions. It felt much less like a typical mentorship the place you’re handed a guidelines and extra like being a part of an actual engineering workforce working by way of a fancy drawback collectively.”

CloudNativePG

The CloudNativePG Chaos Testing challenge validates PostgreSQL cluster resilience on Kubernetes by combining Jepsen consistency testing (mathematical proofs that no knowledge is misplaced or corrupted) with LitmusChaos fault injection (repeatedly killing the first PostgreSQL pod to drive automated failovers).

Mentee: Yash Agarwal (learn Yash’s private weblog publish about mentorship expertise)

Mentors: Gabriele Bartolini, Jonathan Gonzalez, Marco Nenciarini, Francesco Canovai

“It was an ideal studying expertise. My mentors had been very skilled and superb of their work and I learnt and grew rather a lot on this course of. I additionally obtained to contribute to a different CNCF challenge Litmus Chaos due to its dependence on this challenge. The suggestions which the mentors supplied was spot on. The code evaluations finished by Jonathan (considered one of my mentors) was very deep dived. I additionally learnt extra about good and normal coding practices and documentation strategies. General it was a tremendous expertise.”

Full listing of tasks

2025 Time period 3: September – November