AI/ML Engineer

Codemonk; Bengaluru; Hybrid; INR 1,800,000 - 3,200,000; Full time

Job summary

Codemonk is hiring an AI Engineer for its Bengaluru Labs team. You will prototype emerging AI/ML capabilities and push the most promising ideas to production at scale, balancing traditional ML with LLM-based agentic systems.

Ideal candidates have 3+ years of production AI/ML experience, strong Python, and a deep understanding of transformer architectures and Hugging Face. You’ll work across ML, MLOps, and cloud environments with a bias for action.

Qualifications

• 3+ years of AI/ML Engineer experience with production deployments spanning ML models and at least one LLM-based/agentic system.
• Strong Python and software engineering fundamentals (version control, testing, design patterns, code review).
• Solid grounding in classical ML (feature engineering, model selection, hyperparameter tuning, ensemble methods) and transformer architectures, attention, tokenization, embeddings.
• Hands-on with PyTorch, Scikit-learn, and the Hugging Face ecosystem.
• Working knowledge of agentic/RAG frameworks and vector stores (LangChain, Pinecone, Weaviate, Qdrant, FAISS).
• Experience with inference servers (vLLM, TGI, Triton) and quantization techniques (GPTQ, AWQ, GGUF).
• Hands-on with MLOps tools and at least one major cloud (AWS, Azure, or GCP).
• High bias for action, strong communication, and intellectual curiosity.

Responsibilities

• Own the complete ML lifecycle, including data curation, feature engineering, model building, evaluation, deployment, monitoring, and retraining for ML and generative AI systems.
• Build and tune classical ML models (regression, classification, ensemble methods, time series) alongside deep learning systems.
• Create scalable training pipelines, CI/CD, model registry, A/B testing, drift detection, and automated retraining; optimize inference latency, throughput, and cost.
• Fine-tune and deploy LLMs/SLMs using LoRA/QLoRA, PEFT, instruction tuning, and RLHF/DPO; apply quantization and distillation.
• Design agentic systems with RAG pipelines, tool calls, memory, planning loops, and multi-agent orchestration with guardrails.
• Build evaluation frameworks (offline/online) and traditional ML eval; diagnose hallucinations, bias, and drift.
• Track SOTA research and present demos to stakeholders and leadership.

Skills

Python

ML engineering

Transformer models

PyTorch

Scikit-learn

Hugging Face ecosystem

LangChain / LangGraph

Vector stores

MLOps tools

Education

B.Tech./M.Tech in CS/DS or equivalent

Tools

Docker

Kubernetes

MLflow

Weights & Biases

Airflow

Kubeflow

Triton

vLLM

Job description

About The Role

We're hiring an AI Engineer for our Labs team, a fast-moving group that prototypes emerging AI/ML capabilities and takes the most promising ones to production at scale. You'll work across both traditional machine learning and generative AI, turning ideas into working demos within days and hardening them into reliable, production-grade systems. Ideal candidates are equally comfortable building a classical predictive model and architecting an LLM-powered agentic system.

Interview Process: Face-to-Face (F2F) interview in Bengaluru.

• Own the complete ML lifecycle, including data curation, feature engineering, model building, evaluation, deployment, monitoring, and retraining, for both predictive ML and generative AI systems
• Build and tune classical ML models (regression, classification, ensemble methods, time series forecasting) alongside deep learning and transformer-based systems
• Build scalable pipelines for training, CI/CD, model registry, A/B testing, drift detection, and automated retraining; optimize inference for latency, throughput, and cost
• Fine-tune and deploy LLMs/SLMs using LoRA/QLoRA, PEFT, instruction tuning, and preference tuning (RLHF/DPO); apply quantization and distillation
• Design and productionize agentic systems, including RAG pipelines, tool/function calling, memory, planning loops, and multi-agent orchestration, with guardrails and observability
• Build evaluation frameworks (offline and online, LLM-as-judge, red-teaming) and traditional ML evaluation (cross-validation, ROC-AUC, precision/recall); diagnose and mitigate hallucinations, bias, and drift
• Track SOTA research across both ML and GenAI, prototype quickly, and present demos/tech talks to stakeholders and leadership

What We’re Looking For

• 3+ years as an AI/ML Engineer or Applied Scientist, with proven production deployments spanning both traditional ML models and at least one LLM-based or agentic system
• Strong Python and software engineering fundamentals (version control, testing, design patterns, code review)
• Solid grounding in classical ML (feature engineering, model selection, hyperparameter tuning, ensemble methods) and a deep grasp of transformer architectures, attention, tokenization, and embeddings
• Hands‑on with PyTorch, Scikit‑learn, and the Hugging Face ecosystem
• Working knowledge of agentic/RAG frameworks (LangChain, LangGraph, LlamaIndex, CrewAI, or AutoGen) and vector stores (Pinecone, Weaviate, Qdrant, pgvector, or FAISS)
• Experience with inference servers (vLLM, TGI, or Triton) and quantization techniques (GPTQ, AWQ, GGUF)
• Hands‑on with MLOps tools (MLflow, Weights & Biases, Airflow, or Kubeflow), Docker, Kubernetes, GPU workloads, and at least one major cloud (AWS, Azure, or GCP)
• High bias for action, strong communication, and intellectual curiosity

Education

B.Tech/M.Tech in Computer Science, Data Science, AI/ML Engineering, or a related quantitative field, or equivalent practical experience backed by a strong portfolio (open source, publications, or production work).

Nice to Have

Open‑source contributions, multimodal model experience, on‑device SLM deployment, or familiarity with LLM security (OWASP LLM Top 10).

Skills: ml,agentic ai,rag,ai,traditional model,llm