Key Responsibilities:-

- Build and maintain ingestion pipelines from BT's source systems (docs, wikis, ticketing, databases)
- Own ETL/ELT jobs, chunking strategy, embedding generation, and pipeline monitoring
- Implement data quality checks, deduplication, and access-control tagging on ingested content
- Manage schema and version changes as BT's underlying systems evolve
- Implement RAG pipelines, multi-agent orchestration, and prompt/context construction logic
- Tune retrieval quality (hybrid search, reranking, context window management)
- Integrate with the client's chosen LLM providers and manage prompt versioning
- Own latency and cost optimization for inference calls
- Own deployment (cloud/on-prem per BT's constraints), CI/CD, and environment management
- Handle auth/SSO, network policies, and compliance requirements specific to BT (telecom-grade security)
- Set up observabilitylogging, tracing, alertingfor the whole pipeline
- Manage integration with BT's existing enterprise systems (ticketing, CRM, internal tools)
- Build evaluation harnesses (accuracy, faithfulness, retrieval precision/recall) for the context pipeline
- Run regression testing whenever prompts, models, or data sources change
- Own guardrailsPII detection, hallucination checks, content filtering
- Publish quality dashboards/reports used in BT governance reviews

Required Skills:-

- Python and SQL
- ETL/ELT tooling (Airflow, dbt, or equivalent) and document chunking/embedding pipelines
- Vector databases (Pinecone, Weaviate,



pgvector, or similar) and embedding models
- RAG architectures, hybrid search, reranking, and context window management
- Multi-agent orchestration frameworks (LangChain, LangGraph, or similar)
- Prompt engineering and prompt versioning; working knowledge of major LLM APIs (OpenAI, Anthropic, Azure OpenAI, Bedrock)
- Cloud platforms (AWS/Azure/GCP) and on-prem deployment patterns; Docker/Kubernetes
- CI/CD pipelines and setting management
- SSO/OAuth, enterprise auth, and networking/security fundamentals for regulated environments
- Observability tooling (Prometheus, Grafana, OpenTelemetry) and building LLM evaluation harnesses
- Regression testing practices and guardrail/PII-detection tooling
- 4+ years in data/ML engineering or applied AI, ideally with a prior LLM system in production

Nice to Have: Experience in telecom or another regulated industry; knowledge graphs/ontology-based retrieval; LLM cost/latency optimization (caching, batching, model routing); exposure to enterprise ticketing/CRM platforms (ServiceNow, Salesforce, Jira); experience presenting quality metrics to governance stakeholders.

Interested Candidates Kindly share your updated resume directly to @

[email protected] with the below details.

- Candidate name
- Contact No
- Email ID
- Skill
- Current Company
- Total Experience
- Rel Experience
- Current Location
- Preferred Location
- Current CTC
- Exp CTC
- Notice Period

📌 Forward Deployed Engineer (Bengaluru)
🏢 Prodapt Solutions
📍 Bengaluru