Job Opportunities

Agentic AI Architect

Key Responsibilities
Agent Design & Architecture:- Architect multi-agent systems using appropriate orchestration
patterns — planner-solver, autonomous teams, goal decomposition, tool-use chains — selected
based on the problem, not the framework- Design memory architectures for agent interaction,
coordination, and context persistence (short-term, long-term, episodic)- Build reusable tooling,
APIs, and interfaces for agent orchestration that clients can extend and maintain- Evaluate and
select LLMs (commercial and open-source), vector databases, and retrieval systems based on client
requirements, cost, and performance
Production Deployment & Operations (Non-Negotiable):- Deploy agentic AI systems in
cloud-native environments (AWS, GCP, Azure) — containerized, auto-scaling, monitored- Design
and implement agent versioning, rollback, and A/B testing strategies — agents in production must
be operable- Build evaluation and quality monitoring systems for agent outputs — regression test
ing, quality scoring, continuous validation- Implement observability for agent workflows — tracing,
logging, cost tracking, latency monitoring across multi-step agent chains- Design guardrails, safety
mechanisms, and compliance controls appropriate to the client’s industry
Client Engagement:- Translate client business requirements into agent-based technical solutions- Lead architectural discussions, PoCs, and hands-on development- Guide client teams in adopt
ing agent frameworks, best practices, and operational standards- Provide clear documentation,
runbooks, and handoff materials
1
Must-Have Skills
Agentic AI & LLM Systems:- 2+ years hands-on experience designing and deploying multi
agent systems- Strong understanding of agent orchestration patterns — not just framework usage,
but understanding why and when to use each pattern- Proficiency with LLM APIs, RAG archi
tectures, vector stores, tool use, and chaining logic- Hands-on experience with both commercial
(GPT-4, Claude, Gemini) and open-source LLMs (Mistral, LLaMA, etc.)- Prompt engineering,
model selection trade-offs, and context window management- Model-cost profiling and budgeting
—API call optimization, caching strategies, batch vs streaming decisions
Production & Cloud (Non-Negotiable):- Experience deploying AI systems in cloud-native en
vironments — not just prototyping, but production deployment with monitoring- Container orches
tration (Kubernetes/ECS), auto-scaling strategies, and infrastructure-as-code- CI/CD pipelines for
AI systems — automated testing, deployment, and rollback- Observability — distributed tracing,
logging, alerting for multi-step agent workflows- Agent versioning and rollout strategies — the
ability to update agents in production without breaking things
Engineering Foundation:- Strong Python development skills- API design and microservices
architecture- Experience with vector databases (Pinecone, Weaviate, FAISS, Qdrant, or equivalent)- Familiarity with agent orchestration frameworks (LangChain, LangGraph, CrewAI, AutoGen,
Haystack, or equivalent) — but understanding the patterns matters more than knowing every
framework
Operational Mindset:- Designs for reliability and operability, not just functionality- Thinks
about monitoring, cost, and failure modes from day one- Can articulate what happens when an
agent fails, hallucinates, or degrades — and has strategies for each