2 hours ago
Bengaluru, IndiaMid Level
Responsibilities
- Build and optimize enterprise-grade RAG pipelines, including document ingestion, indexing, embeddings, hybrid search, re-ranking, and citation-based generation.
- Implement AI safety and guardrail capabilities such as PII detection, data protection, redaction, prompt-injection prevention, toxicity detection, and misuse detection.
- Design long-term AI memory, knowledge retention, knowledge graph, graph-retrieval, and context-management solutions.
- Create evaluation metrics, golden datasets, testing frameworks, benchmarking processes, and response-quality and faithfulness evaluations.
- Develop scalable production AI services and APIs with attention to latency, throughput, reliability, cost, monitoring, observability, and auditability.
- Design autonomous agents and multi-agent workflows with planning, reasoning, tool calling, memory, orchestration, governance, and monitoring.
- Own scoped features independently, write tests, benchmark implementations, and take AI/ML work from experimentation through production deployment and maintenance.
Requirements
- Experience building and operating ML or AI features in production across experimentation, packaging, serving, monitoring, and maintenance.
- Experience with distributed systems, microservices, event-driven systems using Kafka and NATS, REST APIs, real-time communication, API security, JWT, rate limiting, and resilience patterns.
- Experience building AI agents and multi-agent workflows with LangChain and LangGraph, including planning, tool-use, prompt engineering, context management, memory, and multi-LLM integration.
- Strong Python programming skills with FastAPI and Pydantic, plus experience with spaCy, Sentence Transformers, PyTorch, Hugging Face, ONNX, LoRA/QLoRA fine-tuning, vLLM, and RAGAS.
- Experience designing RAG solutions using document chunking, embeddings, vector databases, Elasticsearch, BM25, hybrid search, and cross-encoder re-ranking.
- Experience with RAGAS, DeepEval, benchmarking, LLM-as-a-Judge, human-in-the-loop evaluation, AI safety, prompt-injection prevention, red teaming, and safety benchmarks.
- Experience with PostgreSQL, pgvector, Redis, Kafka, and Neo4j knowledge graphs.
- Desirable experience with ClickHouse, Docker, Kubernetes, OpenTelemetry, Dynatrace, and MLflow, along with scalable cloud-native infrastructure, observability, and MLOps.
Tech Stack
Apache KafkaClickHouseDockerElasticsearchFastAPIKubernetesMLflowNeo4jPostgreSQLPythonPyTorchRedisspaCy
Categories
About BT
BT Group builds and operates telecom networks and digital services for consumers, enterprises, and public‑sector clients, selling broadband, fixed-line voice, mobile (via EE), and managed network, cloud, and cyber security services. It is headquartered in London and listed on the London Stock Exchange, with American depositary shares on the NYSE. BT also owns Openreach, which manages the UK’s fibre and copper access network used by multiple retail providers.