Jneid Jneid

The 5x Engineer

I'm a 2x, maybe 3x engineer on a good day. The aspiration is 5x. But here's the thing -- being a 1x engineer is itself a moving target. What counted as 1x two years ago is table stakes now. AI code tools rewrote the baseline. Life changes shift what you can give. The goal post moves.

The multiplier isn't about raw code output. It's about range. I've written optimization solvers, ML models, data pipelines, SDKs, Terraform modules, and deployment systems. Every time I thought I was done learning a new layer, the layer below it turned out to be the interesting one. The math led to ML. ML led to pipelines. Pipelines led to deployment. Deployment led to infrastructure. Infra is where I landed.

Journey

• 2025 – Present
  Senior MLOps Engineer
  Autodesk
  Finally doing the thing full time. Hosting services, building platform AI/ML features for distributed compute. Unified control plane, self-service tooling, large-scale training infrastructure. Deployment is the job now.
• 2024 – 2025
  Co-Founder
  slashML · Alchemist Accelerator (Class 37)
  Took the deployment obsession and tried to make it a product. Private-cloud AI deployment platform. Built model hosting, an AI Data Scientist, codebase chat. Alchemist Accelerator. Didn't work out, but the deployment problem was real.
• 2022 – 2024
  Data Scientist
  TD Bank
  Moved to insurance. 50+ ML model lifecycles across multiple clouds. Rewrote and automated the entire data pipeline feeding risk models. Built no-code tools for actuaries in my free time. Same pattern again -- the deployment and pipeline work was what pulled me in, not the modeling.
• 2019 – 2022
  ML Researcher & Engineer
  Hydro Québec
  Got introduced to ML for real. Wrote clustering algorithms that cut simulation compute by 90%, fault detection saving $1M. The projects actually made an impact. Published in IEEE and Energies. But getting things deployed and running reliably -- that was the part that stuck with me.
• 2018 – 2020
  MSc, Electrical Engineering
  McGill University
  FRQNT scholarship. Graduate Excellence Award. This is where the math met ML -- power systems forecasting, optimization under uncertainty.
• 2014 – 2018
  BSc, Electrical & Computer Engineering
  Lebanese American University · Beirut
  100% university scholarship. Constraints, optimization, math equations. The foundation before I knew where it would lead.

What I'm Building Now

At Autodesk, I'm building an AI/ML platform pushing for a unified control plane across all services -- centralization of the stack with self-service as the goal. I'm also an in-house tool builder and maintainer -- co-built an AI code review tool that's better than Copilot and maybe CodeRabbit, and adoption has been wild.

• Developing and maintaining 5+ interconnected Terraform modules for infrastructure deployment
• EKS cluster setup, ArgoCD application deployments, promotion pipelines
• Deploying MCP servers in ECS containers
• Distributed training infrastructure -- data/tensor/pipeline parallelism, checkpointing...
• Monitoring, logging, debugging tooling and developer ergonomics
• GitHub webhook apps
• Performance bottleneck investigation across the full ML systems stack

The Startup Chapter

I co-founded slashML. The pitch: create AI applications in your private cloud. Data never leaves your premises. We got into Alchemist Accelerator (Class 37). We built things:

• Co-launched an AI Data Scientist with secure code execution -- deployed to 200+ users
• Co-developed secure model hosting across multiple clouds -- 100+ Llama models deployed
• Co-built a tool to chat with GitHub codebases
• Co-built a full-stack AI deployment studio with GPU workspaces
• Co-developed a Python SDK unifying AWS, GCP, and Azure APIs
• Technical content reaching 5,000+ monthly readers

We interviewed 300+ ML engineers and 100+ AI managers to validate the market. We benchmarked and fine-tuned specialized LLMs. We built an agent system modeled after Magentic One. We did the work.

And it didn't work out. I'm a failed startup founder. Tried B2B, tried dev tools. Nothing prepares you for this journey -- not the technical skills, not the market research, not the accelerator. The thing that's hard about startups is everything at once. I'd do it differently now, but I wouldn't trade the experience.

Publications