AI Engineer

Company Description

BumbleB is an enterprise software company that specializes in building software as a service (SaaS) and platform as a service (PaaS) solutions in the field of data insights and artificial intelligence (AI). Our goal is to enable businesses to leverage the power of AI and data to drive innovation and make intelligent decisions.

Role Description

This is a full-time role (location Bengaluru) for an AI Engineer at BumbleB. You will work at the intersection of machine learning, linguistics, and systems — training, fine-tuning, and aligning language models that power our products. This role requires a deep understanding of how models learn language, how reinforcement learning shapes model behavior, and how to systematically improve output quality through training-time interventions rather than just inference-time tricks.

What You’ll Do

  • Fine-tune and adapt foundation models for domain-specific enterprise use cases
  • Design and run reinforcement learning from human feedback (RLHF) and preference optimization (DPO, KTO) pipelines
  • Analyze and improve linguistic behavior — coherence, factuality, tone, reasoning chains, and failure modes
  • Build training infrastructure — data curation, annotation pipelines, and experiment tracking
  • Evaluate model quality through both automated benchmarks and human evaluation protocols
  • Research and implement techniques from current ML literature to improve model capabilities

Qualifications

  • Strong foundation in machine learning — loss functions, optimization, regularization, generalization
  • Deep understanding of language model internals — transformer architecture, attention mechanisms, positional encodings, tokenization, and how linguistic capabilities emerge from pretraining
  • Hands-on experience with model fine-tuning — supervised fine-tuning (SFT), LoRA/QLoRA, adapter methods
  • Experience with reinforcement learning for language models — RLHF, reward modeling, PPO, DPO, or similar alignment techniques
  • Understanding of computational linguistics — syntax, semantics, pragmatics, and how models represent and generate natural language
  • Strong programming skills in Python with experience in PyTorch or JAX
  • Familiarity with training infrastructure — distributed training, mixed precision, DeepSpeed, or FSDP
  • Experience with experiment tracking and ML workflow tools (Weights & Biases, MLflow)
  • 5+ years of industry or research experience in machine learning or NLP
  • Ability to work independently
  • Excellent problem-solving and analytical thinking skills
  • Effective written and verbal communication skills
  • Master’s or PhD in Computer Science, Machine Learning, Computational Linguistics, or a related field (or equivalent research experience)

Nice to Have

  • Publications or research contributions in NLP, language modeling, or alignment
  • Experience training models from scratch or contributing to open-source model development
  • Familiarity with synthetic data generation and data augmentation for LLM training
  • Understanding of evaluation science — designing benchmarks, inter-annotator agreement, and human evaluation methodology
  • Experience with model interpretability and mechanistic interpretability techniques
Click here to apply.