Machine Learning Engineer - Research

📋 Description

• • Join Level AI as a Machine Learning Engineer – Research and become a core architect of the next generation of Voice AI. You will sit at the intersection of cutting-edge research and real-world impact, translating raw audio and text into business-changing insights for sales teams around the globe.
• • Own the end-to-end lifecycle of NLP models—from ideation and data acquisition to training, evaluation, deployment, and continuous learning in production. You will design pipelines that handle millions of conversational turns per day, ensuring low-latency inference and rock-solid reliability.
• • Pioneer novel solutions in text classification, entity extraction, abstractive summarization, and generative NLP. Whether the challenge is spotting subtle buying signals in a sales call or generating concise follow-up emails, you will experiment with the latest transformer architectures (BERT, DeBERTa, Flan-T5, GPT, Llama, Gemma, DeepSeek) and invent new ones when required.
• • Collaborate daily with product managers, backend engineers, and UX designers to embed your models seamlessly into Level AI’s platform. You will translate customer pain points into crisp ML problem statements, define evaluation metrics that truly matter to revenue teams, and iterate rapidly based on live A/B tests.
• • Optimize for scale and cost. You will profile GPU utilization, implement distributed training strategies, and apply cutting-edge inference optimizations (quantization, pruning, distillation) so that state-of-the-art models run efficiently in our hybrid cloud environment.
• • Build robust post-deployment monitoring that detects drift, bias, and performance degradation before customers notice. You will design feedback loops that retrain models continuously using newly labeled data, ensuring our AI keeps getting smarter over time.
• • Stay ahead of the research curve. You will attend top-tier conferences (NeurIPS, ICML, ACL), publish when appropriate, and bring fresh ideas back to the team. Internal reading groups and paper clubs are part of our DNA—expect lively debates on attention mechanisms and emergent capabilities.
• • Experiment with multimodal models that fuse audio, text, and visual cues to unlock deeper conversational context. If you have ever wondered how eye-tracking or screen-share data could improve intent prediction, this is the place to test that hypothesis.
• • Contribute to our open-source footprint. We believe in giving back: whether it’s a new tokenizer, a benchmarking dataset, or a lightning-fast inference library, your code can impact the wider AI community.
• • Mentor junior engineers and interns. You will run code reviews, pair program on gnarly tensor bugs, and foster a culture of scientific rigor and creative risk-taking. Leadership here is defined by ideas, not titles.
• • Work in a hybrid setup from our Noida office, collaborating face-to-face two days a week while enjoying focused remote time for deep research. Our Mountain Valley leadership team joins virtually, ensuring you are plugged into Silicon Valley’s latest thinking without leaving home.
• • See your work directly influence product roadmaps and customer success stories. When a Fortune 500 client credits Level AI with increasing their win rate by 20%, you will know your models moved the needle.