Cognizant
GenAI Architect, Consulting Principal
About the Role
We’re seeking a GenAI Architect who blends strategic foresight with hands‑on engineering. This role leads the technical architecture for Agentic AI solutions across the enterprise—evaluating use case feasibility, shaping reference architectures, driving POCs and pilots, and guiding platform adoption. You’ll partner closely with product, engineering, data, and security teams to architect scalable, secure, and cost‑effective GenAI capabilities aligned to business priorities. Candidates should have experience with enterprise GenAI architecture (multi‑cloud preferably AWS, or on‑prem) including model access patterns, orchestration, retrieval‑augmented generation (RAG), vector search, and guardrails.
In this role, you will: Strategy & Architecture
Assess use case feasibility (technical complexity, model fit, data readiness, latency/throughput, security/compliance constraints) and produce a go/no‑go recommendation with solution options.
Define the Agentic AI foundation and architect multi‑agent solutions that enable large‑scale, agent‑driven transformation.
Develop reference architectures and pattern libraries (e.g., RAG with enterprise search, function calling/agents, synthetic data generation, multimodal pipelines).
Design and implement POCs/pilots: data connectors, embeddings pipelines, prompt flows, context engineering, evaluation harnesses, and latency/cost benchmarking.
Build RAG pipelines: chunking, embeddings creation, vector indexing (e.g., Azure AI Search/OpenSearch/pgvector/Pinecone), source of truth tracing (citations).
Collaborate with governance teams to implement guardrails and safety layers: content filtering, jailbreak defense, policy checks, role‑based prompts, function calling constraints.
Integrate GenAI with enterprise systems (APIs, microservices, messaging, identity/authorization).
Drive prompt engineering and promptOps: templates, variables, structured output parsing, context windows management, and hallucination reduction.
Partner with business stakeholders to prioritize use cases and translate requirements into technical designs.
Mentor engineering teams; review solution designs; conduct architecture gates and design authority meetings.
Create roadmaps and migration plans for scaling pilots to production, including cost and performance optimization.
Work Model We believe hybrid work is the way forward. Based on this role’s business requirements, this is a hybrid position requiring regular travel (up to 50%) and presence in client or Cognizant offices on the US East Coast or Central Time Zones. Regardless of your working arrangement, we are here to support a healthy work‑life balance through our various wellbeing programs.
What you need to have to be considered
10+ years in software/data/AI architecture; 2+ years hands‑on with LLMs/GenAI in production or pilots.
Experience with cloud AI stacks in AWS—Bedrock, OpenSearch, Lambda, KMS, Step Functions.
Strong in Python/TypeScript/Java and building LLM apps using frameworks like LangChain, LlamaIndex, Semantic Kernel; experience with orchestration (Agents/Tools/Function Calling).
Deep understanding of RAG design: chunking strategies, embeddings (OpenAI, Cohere, text‑embeddings, BGE), vector DBs (pgvector, Pinecone, Weaviate, Milvus, Azure AI Search), and evaluation metrics.
Preferred experience in security & compliance: OAuth/JWT, RBAC/ABAC, encryption, data masking, PII handling, auditability, Responsible AI.
Familiarity with model ecosystem (GPT, Claude, Gemini, Llama, Mistral, DeepSeek) and trade‑offs (context, cost, latency, licensing).
Excellent communication and stakeholder management; ability to present architecture trade‑offs and influence executive decision‑making.
Work Authorization We will only consider applicants who are legally authorized to work in the United States without company sponsorship (H‑1B, L‑1B, L‑1A, etc.).
Salary and Other Compensation Applications will be accepted until January 15, 2026. The annual salary for this position is between $122,400 - $223,500 depending on experience and qualifications. This position is also eligible for Cognizant’s discretionary annual incentive program, based on performance and subject to the terms of Cognizant’s applicable plans.
Benefits
Medical/Dental/Vision/Life Insurance
Paid holidays plus Paid Time Off
401(k) plan and contributions
Long‑term/Short‑term Disability
Paid Parental Leave
Employee Stock Purchase Plan
#J-18808-Ljbffr
In this role, you will: Strategy & Architecture
Assess use case feasibility (technical complexity, model fit, data readiness, latency/throughput, security/compliance constraints) and produce a go/no‑go recommendation with solution options.
Define the Agentic AI foundation and architect multi‑agent solutions that enable large‑scale, agent‑driven transformation.
Develop reference architectures and pattern libraries (e.g., RAG with enterprise search, function calling/agents, synthetic data generation, multimodal pipelines).
Design and implement POCs/pilots: data connectors, embeddings pipelines, prompt flows, context engineering, evaluation harnesses, and latency/cost benchmarking.
Build RAG pipelines: chunking, embeddings creation, vector indexing (e.g., Azure AI Search/OpenSearch/pgvector/Pinecone), source of truth tracing (citations).
Collaborate with governance teams to implement guardrails and safety layers: content filtering, jailbreak defense, policy checks, role‑based prompts, function calling constraints.
Integrate GenAI with enterprise systems (APIs, microservices, messaging, identity/authorization).
Drive prompt engineering and promptOps: templates, variables, structured output parsing, context windows management, and hallucination reduction.
Partner with business stakeholders to prioritize use cases and translate requirements into technical designs.
Mentor engineering teams; review solution designs; conduct architecture gates and design authority meetings.
Create roadmaps and migration plans for scaling pilots to production, including cost and performance optimization.
Work Model We believe hybrid work is the way forward. Based on this role’s business requirements, this is a hybrid position requiring regular travel (up to 50%) and presence in client or Cognizant offices on the US East Coast or Central Time Zones. Regardless of your working arrangement, we are here to support a healthy work‑life balance through our various wellbeing programs.
What you need to have to be considered
10+ years in software/data/AI architecture; 2+ years hands‑on with LLMs/GenAI in production or pilots.
Experience with cloud AI stacks in AWS—Bedrock, OpenSearch, Lambda, KMS, Step Functions.
Strong in Python/TypeScript/Java and building LLM apps using frameworks like LangChain, LlamaIndex, Semantic Kernel; experience with orchestration (Agents/Tools/Function Calling).
Deep understanding of RAG design: chunking strategies, embeddings (OpenAI, Cohere, text‑embeddings, BGE), vector DBs (pgvector, Pinecone, Weaviate, Milvus, Azure AI Search), and evaluation metrics.
Preferred experience in security & compliance: OAuth/JWT, RBAC/ABAC, encryption, data masking, PII handling, auditability, Responsible AI.
Familiarity with model ecosystem (GPT, Claude, Gemini, Llama, Mistral, DeepSeek) and trade‑offs (context, cost, latency, licensing).
Excellent communication and stakeholder management; ability to present architecture trade‑offs and influence executive decision‑making.
Work Authorization We will only consider applicants who are legally authorized to work in the United States without company sponsorship (H‑1B, L‑1B, L‑1A, etc.).
Salary and Other Compensation Applications will be accepted until January 15, 2026. The annual salary for this position is between $122,400 - $223,500 depending on experience and qualifications. This position is also eligible for Cognizant’s discretionary annual incentive program, based on performance and subject to the terms of Cognizant’s applicable plans.
Benefits
Medical/Dental/Vision/Life Insurance
Paid holidays plus Paid Time Off
401(k) plan and contributions
Long‑term/Short‑term Disability
Paid Parental Leave
Employee Stock Purchase Plan
#J-18808-Ljbffr