Ethical Den is hiring an AI Engineer specialising in RAG and AI Agents to work on Hiyvr Technologies.
Every Hiyvr customer may have different products, prices, policies, FAQs, scripts, CRM data and operational information. Our voice agents must retrieve the correct information quickly and perform approved business actions without confusing data between customers.
Because Hiyvr operates through live telephone conversations, retrieval speed is extremely important.
A RAG pipeline that takes several seconds may work for a chatbot, but it is not acceptable for a real-time voice agent.
KEY RESPONSIBILITIES
• Build multi-tenant RAG architecture.
• Develop document ingestion and indexing pipelines.
• Design chunking strategies.
• Build embedding pipelines.
• Implement semantic and hybrid search.
• Evaluate and work with vector databases.
• Implement metadata filtering and customer-level data isolation.
• Build retrieval evaluation datasets.
• Improve retrieval precision and recall.
• Develop caching strategies for frequently requested information.
• Reduce retrieval latency for live voice interactions.
• Build tool-calling workflows.
• Integrate CRM, appointment, lead-management and business systems.
• Implement API integrations for business actions.
• Build validation, retries and duplicate-action prevention.
• Develop fallback behaviour when tools or external systems fail.
• Build multi-step AI agent workflows.
• Measure retrieval latency at p50, p95 and p99.
• Work closely with the Senior AI Engineer and backend team.
• Develop testing and regression frameworks for RAG and agent behaviour.
REQUIRED SKILLS
• Python
• Retrieval-Augmented Generation
• Embeddings
• Vector databases
• SQL
• APIs
• JSON
• Redis or similar caching systems
• Document processing
• Semantic search
• LLM tool calling
• AI agent workflows
• Linux
• Git
EXPERIENCE WITH ONE OR MORE OF THE FOLLOWING IS PREFERRED
• Qdrant
• Milvus
• FAISS
• pgvector
• Weaviate
• Elasticsearch
• OpenSearch
• LangGraph
• LlamaIndex
• LangChain
Framework experience is helpful, but understanding the underlying architecture is more important than knowledge of a specific framework.
VALUABLE TO HAVE
• Multi-tenant SaaS experience
• Real-time systems
• Voice AI
• CRM integrations
• Fast retrieval systems
• Reranking
• Knowledge graphs
• Multilingual embeddings
• Bengali/Hindi document retrieval
EDUCATIONAL QUALIFICATION
Preferred
B.Tech/B.E./M.Tech/MCA/M.Sc. in Computer Science, Artificial Intelligence, Machine Learning or a related field.
Strong candidates with approximately 1.5 to 2 years of highly relevant production RAG experience may also be considered.
WHAT WE WILL EVALUATE
Candidates should be able to explain:
• Why RAG systems retrieve incorrect information
• How to diagnose retrieval failures
• How to reduce retrieval latency
• How to safely execute business actions
• How to isolate data between customers
HOW TO APPLY
Please apply with
• Updated CV
• Current location
• Current CTC
• Expected CTC
• Notice period
• GitHub / portfolio / project links
• Short description of the most relevant RAG or AI-agent system you have built
ABOUT HIYVR TECHNOLOGIES
Hiyvr Technologies is developing real-time AI voice agents for business conversations, with a strong focus on Indian languages, natural speech, low latency and practical business workflows.
Hiyvr is being developed by Ethical Den. This is a full time, on-site role based in Kolkata.
📌 AI Engineer - RAG & AI Agents (Kolkata)
🏢 Ethical Den
📍 Kolkata
Reply to this offer
Impress this employer describing Your skills and abilities, fill out the form below and leave Your personal touch in the presentation letter.