2025—Now
Senior AI/ML Engineer
Labelbox
Translating frontier-lab requirements into production RL data and environments; evaluating capability gaps across terminal, software, and cognitive tasks.
RL gymsEvaluationRLHF / RLAIF
01 / Research paths
Follow the model lifecycle.
Three connected bodies of work: how models learn, how they improve through feedback, and how they run reliably in production.
01
Foundations
Pre-trained
Data, tokenization, model architecture, and the systems that create foundation models.
Coming soon ↗ 02TrainRL
Post-trained
Reinforcement learning, environments, graders, evaluations, and specialist agents.
Read the public post-training blog ↗ 03Serving
Inference
Paged KV caching, continuous batching, fused kernels, and the systems work behind reliable low-latency serving.
Coming soon ↗02 / Experience
Building across the whole AI stack.
Research, data, infrastructure, deployment—the hard problems rarely respect team boundaries.
2024—2025
Principal AI Developer
HTC Inc
Built hot-swap model adapters that cut deployment time by 80% and architected an end-to-end AI workflow management system.
Model adaptersFine-tuningMLOps
2023—2024
Senior Applied AI Scientist
Usable Machines
Built on the WhiteRabbitNeo foundation model and reduced task perplexity from 13 to 5 through preference optimization and task-specific fine-tuning.
Foundation modelsRLHF / RLAIFFine-tuning
2023
Adjunct Lecturer, NLP
Northwestern University
Created a build-an-LLM course and researched controlled legal-text generation with T5, reaching a ROUGE-L F1 score of 0.44.
NLPTeachingT5
2020—2021
Principal AI Developer
Code Techniq
Led an eight-person engineering team, built browser-based vision systems, and designed NLP pipelines for knowledge and medical applications.
BERTTensorFlow.jsML systems
03 / Point of view
The interesting work begins where the demo ends.
I care about the full learning loop: the data we choose, the worlds we ask models to navigate, the failures we measure, and the systems that carry those lessons into production.
Research should leave the notebook.
— VK
04 / Ideas from the workbench
Ideas from the workbench
Notes on alignment, RL environments, and the signals that help language models improve through experience.
Post-training is a stack.
A map of supervised tuning, preferences, reward models, RL, safety, agent trajectories, distillation, and continual improvement.
Read the current blog ↗Current research focus
What the first release maps
- 01 Supervised fine-tuning
- 02 Preference optimization
- 03 Reinforcement learning
- 04 Evaluation and safety
01 The shared structure
Release the research at the pace it can be tested.
The full pre-training, post-training, and inference libraries are preserved. For now, only the foundational post-training guide is public; future field notes will appear one at a time through the research feed.
05 / About
Researcher’s curiosity.
Engineer’s discipline.
I’m Vikram Kharvi, an applied AI researcher and engineer in California. I’ve worked across the AI lifecycle—from training and preference optimization to evaluation, deployment, and monitoring.
I hold a Master of Artificial Intelligence from Northwestern University and a Bachelor of Information Science Engineering from PES University.
EducationNorthwestern UniversityMS, Artificial Intelligence · 3.8/4
RecognitionHack2Innovate WinnerSamsung · NVIDIA · PWC · T-Hub
CertificationNVIDIA Agentic AICertified Professional
PYTORCH✦TRANSFORMERS✦RLHF✦RL ENVIRONMENTS✦AGENT EVALUATION✦KUBERNETES✦PYTHON✦MODEL EVALUATION✦
PYTORCH✦TRANSFORMERS✦RLHF✦RL ENVIRONMENTS✦AGENT EVALUATION✦KUBERNETES✦PYTHON✦MODEL EVALUATION✦
06 / Research feed
Follow the
learning loop.
Subscribe for new field notes on reinforcement learning, LLM post-training, agent evaluation, and efficient inference systems.
Subscribe via RSS Works with Feedly, Inoreader, NetNewsWire, and other feed readers.