About the role
As a Senior AI Research Engineer at TheWebVale, you will investigate modern advancements in machine learning and language models. You will explore small language models, reasoning distillation, vision-language systems, and quantization techniques to unlock practical advantages for client products.
You will partner directly with our founder to translate research breakthroughs into production prototypes in two-to-three week cycles. You will write rigorous technical analyses, benchmark open models, and guide studio technology adoption.
What you will do
- Evaluate newly released open-weight models including Llama, Mistral, and Qwen architectures.
- Benchmark model speed, VRAM footprint, and reasoning accuracy across hardware setups.
- Implement quantization techniques (AWQ, GGUF, EXL2) to run models on cost-effective servers.
- Develop working proof-of-concept prototypes to validate novel AI workflows for clients.
- Author concise technical briefs detailing performance characteristics and deployment costs.
- Mentor engineering teammates on model fine-tuning fundamentals and prompt evaluation.
What you will need
- Have 5+ years of software engineering or applied machine learning research experience.
- Demonstrate deep understanding of transformer architectures, attention mechanisms, and weights.
- Show strong hands-on proficiency in Python, PyTorch, Hugging Face Transformers, and CUDA.
- Exhibit practical experience running local inference with llama.cpp, vLLM, or Ollama.
- Possess a strong scientific mindset with disciplined benchmarking and documentation habits.
- Explain complex machine learning discoveries in accessible terms to product builders.
Nice to have
- Authored research papers, technical blogs, or open-source machine learning contributions.
- Experience fine-tuning models on domain-specific datasets using LoRA or QLoRA.
- Familiarity with multimodal vision-language models and audio transcription models.