the anti job boardactive 4mo ago
Anant Goel

Anant Goel

Software Engineer @ Abnormal AI

Vancouver, BC6+ yearsflexible2 weeks notice

Summary

Anant Goel is an experienced ML Engineer with over 6 years of expertise in building and deploying machine learning systems, particularly in the realm of large language models. He has a strong background in reinforcement learning and has successfully contributed to significant projects at Microsoft and Abnormal AI, showcasing his ability to innovate and optimize complex systems.

Looking for

Senior Software EngineerData Scientist / ML Engineer

Skills

PythonGoC++C#SQLTypeScriptPyTorchTensorFlowscikit-learnspaCyLLM fine-tuningreinforcement learningRAG pipelinesquantizationprompt engineeringmodel evaluationLightGBMDNNsLSTMsAzureAWSModalDockerCI/CDGPU optimizationdistributed inferencemicroservicesREST APIs

Preferences

Open to

remote

#mlengineer#6+years#softwareengineering#machinelearning#ai#reinforcementlearning#llmfine-tuning#cloudinfrastructure#python#go#c++#c##sql#typescript

The story

What Anant is looking for

The day-to-day I'm happiest in is somewhere between research and product, where the two aren't cleanly separated. I'm at my best when I can take an ambiguous problem and work it from hypothesis through experiment to something a customer actually uses. In my previous roles, I've usually been the bridge between research ideas and production needs. This has given me experience to operate on both ends and I quite enjoy that.

Why Anant is exploring

Abnormal has grown, and that's a good thing but it means the kind of bets I want to make aren't the right bets for a company at that stage. I want to go back to smaller teams where I can build 0→1 products again and take product risks that would be irresponsible in a mature product. The next chapter should be a place where the research is ambitious and the path to production is short.

What won't fit on a resume

I've been consistently lucky to work with teams that "live in the future" and work on problems before there is a playbook. At Microsoft I had early access to OpenAI models and was solving LLM problems that didn't have established solutions yet. We were experimenting with things like context chunking strategies when models were still stuck at 8k context. We were fine-tuning models before qLORA was public and made it accessible. At Abnormal now, I'm defining autonomous agents that operate directly in customer inboxes - a year ago, before the agentic AI wave really hit, that wasn't on anyone's roadmap. I gravitate toward problems where the right approach isn't obvious yet.

The Anti Job Board

Skip the application black hole. Get warm intros to founders hiring now.

Create your profile