Application — CERN IT-CD-PI-2026-63-LD · Machine Learning Engineer

I Build the ML Platforms That Turn Accelerator Data Into Discovery.

5+ years shipping production AI/ML services — model training, serving, and LLM/agent systems — with a prior CERN collaboration on Rucio. Ready to bring enterprise-scale MLOps and open-source depth to CERN's next-generation ML platform.

Prior CERN Collaborator — GSoC 2022, Rucio / ATLAS / CMS
Why I Fit CERN Explore Projects Download Resume

0

Years in production AI/ML engineering

0

Users impacted by shipped AI tooling

0

Cost reduction on optimized inference

0

Models fine-tuned & quantized for production

About

Enterprise-hardened ML engineering meets scientific computing roots.

My career bridges two worlds: I've built patent-pending AI systems at JPMorganChase under strict enterprise constraints, while also contributing to CERN's open-source ecosystem through Google Summer of Code (Rucio, 2022). That dual perspective — shipping at scale and collaborating in open-source scientific communities — is exactly what this role demands.

Train → Optimize → Serve

Full ownership from model training and HPO through quantization (ONNX/AWQ/GPTQ) to GPU-optimized inference at 100+ tok/s.

Ship & Operate at Scale

Kubernetes-native deployments, CI/CD pipelines, observability dashboards, and release governance for 250K+ user workloads.

LLM & Agent Systems

Enterprise RAG/CAG architectures, agentic frameworks, evaluation suites, and drift controls — patent-pending work.

Document & Align

Reference architectures, unified documentation (Google Season of Docs), and cross-team collaboration with researchers and engineers.

Why CERN, Why Me

Every responsibility mapped to proven experience.

Develop and operate the ML service — training, HPO, and serving on accelerator resources 5+ yrs production ML · SageMaker · 100+ tok/s inference · GPU optimization
Integrate LLM and AI agent capabilities — scalable serving, agent frameworks, multi-tenant access Patent-pending agentic AI platform · 250K+ users · RAG/CAG · secure access patterns
Extend MLOps — deployment, observability, lifecycle management of models and RAG pipelines Full MLOps ownership · CI/CD · Kubernetes · LLM eval suite · drift monitoring
Reference architectures, documentation, and best practices for users Google Season of Docs · Creative Commons docs unification · architecture leadership
Collaboration with IT groups, departments, and external research/industry orgs CERN GSoC (Rucio) · ATLAS/CMS stakeholders · open-source contributor · cross-team delivery

Featured Work

Selected projects in ML platforms, scientific computing, and model operations.

Rucio WebUI (CERN GSoC)

LHC-scale data workflows

Re-architected the Rucio WebUI with CERN collaborators to improve scientific data management usability for ATLAS and CMS experiments, handling petabyte-scale data movement and long-lived operational workflows.

  • Open Source
  • Rucio
  • CERN
  • ATLAS/CMS
  • Scientific Data Mgmt

Enterprise Agentic AI Platform

250K+ users impacted

Built a patent-pending multimodal language agent framework with LLM/RAG/CAG patterns, secure multi-tenant access, and productionized services for internal workflows.

  • Python
  • LLM/RAG/CAG
  • AWS SageMaker
  • Multi-Tenant
  • MLOps

Inference Optimization Pipeline

100+ tokens/sec at 128K+

Fine-tuned and quantized multiple LLM/SLM models with ONNX, AWQ, and GPTQ, deploying GPU-optimized inference services for large-context, high-throughput usage.

  • ONNX
  • AWQ
  • GPTQ
  • Kubernetes
  • GPU Optimization

LLM Evaluation & Benchmarking Suite

Drift and hallucination controls

Delivered an evaluation dashboard with LLM-as-a-Critic, self-reflection, and black-box methods for adaptive thresholding, quality tracking, and lifecycle governance.

  • LLM Evals
  • Observability
  • Adaptive Thresholding
  • Lifecycle Mgmt

Technical Skills

Core competencies aligned to the role.

ML Systems & Model Serving
LLM / Agent Frameworks
MLOps & CI/CD
Kubernetes & Cloud Native
GPU / Accelerator Platforms
Open Source Ecosystems
Architecture & Documentation
Collaboration & Communication

Tools & Technologies

PythonPyTorchONNX Runtime LangChainLlamaIndexHuggingFace DockerKubernetesAWS SageMaker CI/CDTerraformPrometheus TypeScriptSQL/NoSQLVectorDB

Experience

5+ years across enterprise AI/ML and open-source scientific computing.

  1. Senior AI Research Engineer — JPMorganChase

    Mar 2021 – Present · Bengaluru, India

    Progression from core engineering to senior AI platform ownership:

    • Software Engineer I (1 year 4 months)
    • Software Engineer II (2 years 2 months)
    • Senior Research Software Engineer (SDE III)
    • Senior AI Research Engineer (current)

    Led delivery of multimodal agentic AI systems, LLM evaluation frameworks, and GPU-optimized inference pipelines, including serving reliability, lifecycle observability, and release controls.

  2. Research Collaborator (Google Summer of Code) — Rucio, CERN

    Jun 2022 – Oct 2022 · Geneva, Switzerland (Remote)

    Contributed core features to the Rucio WebUI — the scientific data management system used by LHC experiments including ATLAS and CMS. Collaborated directly with CERN scientists and engineers on re-architecting web interface components for petabyte-scale data workflows. Worked within CERN's open-source development model, participating in code reviews, community discussions, and upstream contribution practices.

  3. Technical Writer (Google Season of Docs) — Creative Commons

    Jun 2020 – Mar 2021 · Mountain View, CA (Remote)

    Authored and consolidated developer documentation across multiple open-source repositories (Vocabulary, Vue-Vocabulary, Fonts), improving onboarding and contributor productivity with a unified documentation site.

Education

M.S. Computer Science

Georgia Institute of Technology

2025

B.E. Information Science & Engineering

Ramaiah Institute of Technology

2017–2021 · CGPA 9.6/10 · Best Outgoing Student

Languages

English Fluent French Beginner Hindi Native Marathi Native Kannada Beginner

Publications, Talks & Credentials

Research, technical documentation, and recognition.

Publication

Spider Monkey Optimisation Algorithm — Springer

Book chapter published in the "Nature Inspired Computing for Data Science" volume series.

Publication (In Progress)

AgentVaccine: Third-Agent Intervention for Safer Autonomous Multi-Agent Systems

ACM CCS Conference — Studying third-agent intervention to reduce harmful influence in LLM-based agent conversations.

Poster Presentation

DocAid: AI-Powered Medical Documentation Assistant

IEEE CCEM Poster — Intelligent clinical note generation system reducing physician documentation burden.

Google Season of Docs

Creative Commons Documentation Unification

Authored implementation guides and unified docs for Vocabulary, Vue-Vocabulary, and Fonts.

Certifications

AWS Certified (Developer Associate, Cloud Practitioner)

Validated cloud and deployment foundations for building scalable, production AI/ML systems.

Recognition

Software Engineer of the Quarter — JPMorganChase

Awarded for excellence in delivery and leadership while shipping high-impact AI capabilities.

nimishnb98@gmail.com LinkedIn GitHub Hugging Face