HY
Lead AI Engineer @ Best Buy India · open to frontier-AI roles & collaboration

I build the layer between
"it works in a notebook"
and "it serves real traffic."

Production engineer for LLM fine-tuning

I'm Hari — a production LLM/ML systems engineer with 10+ years shipping models at scale. Today I lead AI engineering at Best Buy India: LLM fine-tuning (LoRA/QLoRA via Unsloth/PEFT, multi-GPU) and serving (vLLM, GKE, hot-swappable adapters), plus real-time fraud detection at 40k events/sec. Before that, GPU-accelerated cybersecurity ML on 8×A100 DGX with NVIDIA Morpheus, supply-chain optimization, and computer vision for semiconductor QA. My roots run all the way back to enterprise mainframes — so I know how real systems stay up.

10+ yrs
Production ML & data
2 Clouds
GCP & Azure, deep
Lead
AI Engineer @ Best Buy
End→End
Data ➜ Model ➜ MLOps
// impact

Proof, not adjectives

Selected outcomes from production work. Numbers reflect real systems I've built, trained, and shipped.

$300K saved
Self-hosted open-source LLMs (LoRA fine-tune → vLLM serve) — stronger self-reliance & data governance.
40k/sec
Real-time fraud detection throughput in production.
8×A100
GPU-accelerated cybersecurity ML on DGX with NVIDIA Morpheus.
$2.5M saved
Azure ML quality prediction — warehouse trucks inspected/day 130→160, +17% defect detection.
Cybersecurity case volume vs. prior pipeline — threats automatically routed to the right team.
Billion-row
Datasets engineered & modeled on GPU (RAPIDS cuDF / CuPy).
// about

From mainframes to frontier-scale AI

A rare engineer who has worked both ends of the stack — the COBOL/Db2 systems that quietly run enterprises, and the LLM systems reinventing them.

I'm a production LLM/ML systems engineer. My focus is the unglamorous, decisive part: taking a model from "works in a notebook" to "serves real traffic" — reliably, observably, and at scale.

I started on enterprise mainframes — COBOL, JCL, Db2 — where reliability isn't optional. That foundation became my edge: I moved through data science, GPU-accelerated ML (8×A100 DGX, NVIDIA Morpheus), and into LLM engineering — fine-tuning with LoRA/QLoRA via Unsloth/PEFT and serving with vLLM and hot-swappable adapters.

Today I lead AI engineering at Best Buy India: real-time fraud detection at 40k events/sec, self-hosted LLM initiatives, and the MLOps that keeps it all alive across GCP and Azure.

I'm now going deeper — toward inference internals and frontier-scale AI: turning "framework user" into "framework builder." I build in public, write what I learn, and chase the hard problems.

NOW · Best Buy India
Lead AI Engineer
LLM fine-tuning (LoRA/QLoRA, Unsloth/PEFT) & serving (vLLM, GKE, hot-swap adapters); real-time fraud detection @ 40k events/sec.
GPU & Security ML
GPU-Accelerated ML Engineer
8×A100 DGX with NVIDIA Morpheus; anomaly detection on billion-row datasets via RAPIDS.
Optimization & CV
Data Scientist
Azure ML quality prediction (130→160 trucks inspected/day, $2.5M saved) & OR-Tools route optimization; computer vision for semiconductor QA; forecasting & BI.
Foundations
Enterprise Systems · Mainframe
COBOL, JCL, Db2 — scale, reliability, and how businesses really run.
// stack

The tools, end to end

Leading with the LLM systems stack — then the full lifecycle that gets a model into production and keeps it there.

🧠 LLM Engineering

LLM Fine-TuningLoRA / QLoRAPEFT UnslothvLLMHot-Swap Adapters Generative AIRAGEval & PromptingNLP

ML at Scale

Real-Time MLMulti-GPU / Distributed8×A100 DGX NVIDIA MorpheusRAPIDS cuDFCuPy ForecastingAnomaly DetectionRecommendation Computer VisionOptimization · OR-ToolsCUDA-adjacent

🐍 Languages & Frameworks

PythonPyTorchHF Transformers scikit-learnPandasTensorFlow / Keras FastAPISQLShell COBOL · JCL (heritage)

☁️ Cloud · MLOps · Infra

GCP — Vertex AIGKEDataflowPub/Sub Azure MLKubernetesDocker MLflowKubeflowAirflow Seldon CoreCI/CDGrafana · Prometheus ELK StackPower BIMS SQL · Db2
// work

Building in public

Open-source work, from local-first AI tooling to production-minded utilities — plus the deeper systems I'm building now.

// words

I write & I solve

Sharing what I learn, and keeping the fundamentals sharp.

Let's build something intelligent.

Open to senior AI / ML / research-engineering roles and ambitious collaborations. If you're pushing the frontier — let's talk.