AI Architect

Insight Global; India; Hybrid; INR 3,330,000 - 4,070,000; Full time

Job summary

A leading life sciences company is seeking an AI Architect in India to design and deploy Generative AI solutions. The role requires expertise in AI governance, LLMs, and collaboration with cross-functional teams in a hybrid work environment. Ideal candidates should have extensive experience in AI solution architecture, particularly with LLM applications and data governance. Competitive compensation up to 37 LPA offered.

Qualifications

• 7+ years of experience in AI/ML solution architecture with LLM focus.
• Success in deploying production LLM applications.
• Hands-on with LLM APIs and orchestration frameworks.

Responsibilities

• Lead design and deployment of Generative AI solutions.
• Collaborate to build end-to-end LLM workflows.
• Utilize Databricks for data governance.

Skills

AI/ML solution architecture

LLMs and Generative AI

Python for orchestration pipelines

Databricks ecosystem

Data governance

Vector databases

Cloud or hybrid environments

Communication skills

Job description

Overview

A global life sciences company is seeking an AI Architect to join the team in India (hybrid onsite OR remote). This high-impact role is ideal for an experienced architect passionate about applying cutting-edge LLM technology to solve complex problems in healthcare, research, and regulated environments. The architect will lead the design and deployment of scalable, secure, and compliant Generative AI solutions that power use cases such as domain-specific assistants, semantic search, summarization, and intelligent knowledge retrieval. They will collaborate with data engineering, platform, and compliance teams to build end-to-end LLM workflows—integrating RAG pipelines, vector stores, prompt orchestration, and real-time inference. They will use Databricks Unity Catalog to ensure robust governance of training and inference datasets, metadata, lineage, and access controls. The role requires a strong architectural mindset, a deep understanding of AI governance in regulated environments, and the ability to work closely with scientific and business stakeholders to translate domain needs into effective AI systems.

This role can pay up to 37 LPA.

Responsibilities

• Lead the design and deployment of scalable, secure, and compliant Generative AI solutions for use cases such as domain-specific assistants, semantic search, summarization, and intelligent knowledge retrieval.
• Collaborate with data engineering, platform, and compliance teams to build end-to-end LLM workflows, including RAG pipelines, vector stores, prompt orchestration, and real-time inference.
• Utilize Databricks Unity Catalog to govern training and inference datasets, metadata, lineage, and access controls.
• Translate domain problems into modular, scalable, and secure AI architectures; work with scientific and business stakeholders to define requirements and deliverables.

Qualifications

• 7+ years of experience in AI/ML solution architecture, including 2+ years working directly with LLMs and Generative AI
• Proven success deploying production LLM applications, including prompt engineering, fine-tuning, and retrieval-augmented generation
• Hands-on experience with enterprise-grade LLM APIs (e.g., OpenAI, Claude, Gemini) and orchestration frameworks (e.g., LangChain, LlamaIndex)
• Proficient in Python for building orchestration pipelines, APIs, and experimentation frameworks
• Strong expertise with Databricks ecosystem—Unity Catalog, Delta Lake, MLflow, and cluster orchestration
• Deep knowledge of data governance, metadata management, and data lineage in large-scale, regulated environments
• Experience integrating vector databases (FAISS, Weaviate, Pinecone) and implementing feedback systems (e.g., RLHF, user feedback loops)
• Ability to map domain problems to AI workflows in a modular, scalable, and secure way
• Familiarity with deploying models in cloud or hybrid environments (AWS, Azure, or GCP)
• Clear and effective communicator, capable of working cross-functionally with technical and non-technical teams