Talha Amjad

Work

Work History.

A selection of work across product delivery, AI reasoning, cloud-native infrastructure, and mentoring.

Sep 2026 – Present

LFX Mentee - Generic xPU Topology-Aware Scheduling

The Linux Foundation Mentorship · CNCF Volcano · Remote

As an LFX mentee with CNCF Volcano, I am designing and implementing a vendor-neutral scheduler-side framework that makes physical xPU topology and concrete device availability visible during scheduling.

  • Model device, domain, and fabric availability through a provider abstraction that normalizes annotations, CRDs, Device Plugin companion data, vendor APIs, and DRA ResourceSlices.
  • Design topology-aware filtering, scoring, and gang-wide planning with atomic reservation and rollback of nodes, device domains, and concrete device IDs before binding.
  • Cover HCCS, NVLink, NVSwitch, multi-node, and rack-scale fabric layouts, with DRA and Device Plugin enforcement plus KWOK-based functional and scale validation.
GoKubernetesVolcanoxPU topologyDRADevice PluginKWOK

Aug 2026 – Present

Community Member / Contributor

Volcano (A Cloud Native Batch System) · Remote

Contribute scheduler and infrastructure fixes across device-aware reclaim and preemption, GPU and Pod resource accounting, DRA scheduling, controller reliability, and queue-scoped admission design.

  • Improve reclaim and preemption paths where device feasibility must remain correct after tentative evictions.
  • Work on GPU and Pod resource accounting, DRA scheduling behavior, and controller-level reliability fixes.
  • Shape queue-scoped admission controls that add local policy without weakening cluster-wide safety boundaries.
GoKubernetesVolcanoSchedulingDRAAdmission Webhooks

Oct 2025 – Jan 2026

AI Reasoning Engineer

Turing · Palo Alto, CA (Remote)

Designed complex mathematical and algorithmic challenges to train large language models. Developed production-grade C/C++ solutions with detailed time and space complexity analysis.

  • Designed mathematical and algorithmic tasks intended to evaluate and improve model reasoning.
  • Produced production-grade C/C++ reference solutions for the challenges.
  • Documented time and space complexity so each solution could be evaluated with clear technical criteria.
C/C++AlgorithmsMathematicsLLM trainingComplexity analysis

Sep 2024 – Oct 2025

Associate Software Engineer

Maima Soft · Lahore, Pakistan (Hybrid)

Promoted from intern to associate in under one year. Built and maintained enterprise accounting, pharmacy, and school management systems serving thousands of users.

  • Progressed from intern to associate engineer in under one year through client-facing product work.
  • Built and maintained accounting, pharmacy, and school-management systems used in daily operations.
  • Worked across frontend, backend, and data concerns to keep enterprise workflows dependable for users.
React.NETASP.NETSQL ServerEnterprise systems

Jul 2024 – Sep 2024

Intern Software Engineer

Maima Soft · Lahore, Pakistan (Hybrid)

Worked alongside senior engineers on client-facing projects, gaining hands-on experience with .NET, SQL Server, and React in production.

  • Worked alongside senior engineers on client-facing software delivery.
  • Built practical experience with .NET services, SQL Server data work, and React interfaces.
  • Learned how production application work moves from implementation to real client use.
React.NETSQL ServerClient delivery

2025 – Present

Trainer / Mentor

Dev Weekends · Lahore, Pakistan (Hybrid)

Mentor students and junior developers on problem-solving, data structures, and full-stack development.

  • Mentor students and junior developers through structured problem-solving practice.
  • Teach data structures and algorithms as practical tools for building stronger engineering fundamentals.
  • Guide full-stack development work with an emphasis on clear implementation and steady learning.
Problem solvingData structuresAlgorithmsFull-stack development