Senior Agentic AI Solution Architect

ThinkWise Consulting LLP; Hyderabad; Remote; INR 3,000,000 - 6,000,000; Full time

Job summary

ThinkWise Consulting LLP is seeking a Senior Agentic AI Solution Architect to design an auditable, safe multi-agent platform. You will own architecture across ingestion, reasoning, and governance for an enterprise-grade system, collaborating with senior client leadership.

You will lead architecture decisions, define knowledge layers, and drive end-to-end pilots while ensuring compliance and safety in a regulated domain.

Qualifications

• 10+ years in software engineering/architecture, with 3+ years hands-on with agentic or multi-agent LLMs in production.
• Proven greenfield architecture ownership on a complex, ambiguous system.
• Deep multi-agent orchestration: planning, reasoning loops, memory, and collaboration.
• Strong RAG/knowledge architecture: embedding, hybrid/semantic retrieval, vector stores, grounding.
• Hands-on with agentic frameworks (LangGraph/LangChain) and MCP or comparable tool layers.
• Demonstrated AI safety and evaluation work: guardrails and evaluation harnesses.
• Strong Python and/or Java, API-first and microservices background.
• Exceptional client-facing communication with senior technical executives.

Responsibilities

• Own end-to-end architecture: ingestion, knowledge/retrieval, agent reasoning, tool access, governance.
• Design knowledge lake across telemetry, VoC signals, CRM, backlog, documents, repositories.
• Architect MCP gateway, registry, servers, and connectors for governed tool/data access.
• Design multi-agent reasoning core: planner/re-planner, loops, memory, collaboration.
• Architect solutions using LLMs, RAG, knowledge systems, vector databases, and semantic search.
• Build governance as a first-class layer: audit trails, grounding, guardrails, human-in-the-loop, graceful degradation.
• Establish AI governance, security, explainability, observability, and Responsible AI standards.
• Define evaluation harness, quality metrics, and measurable baselines.
• Drive end-to-end pilot on a real-use case before scaling; set offshore delivery direction.
• Create blueprints, reference implementations, and reusable frameworks; guide engineering teams.

Skills

Agentic LLM systems

Multi-agent orchestration

Python/Java

API-first architecture

Microservices

RAG and knowledge architecture

LangGraph

LangChain

Tool access governance

Education

Bachelor's or Master's in CS/Engineering/AI

Tools

LangGraph

LangChain

MCP gateway

Vector stores

Job description

Job Location: Remote — Hyderabad, Indore . Ahmedabad

Experience: 10+ Years

Role Type: Hands-on technical architect — not a delivery or engagement management role

About the Role:

We are looking for an experienced Senior Agentic AI Solution Architect to serve as technical design

partner for a strategic enterprise client building a greenfield, production-grade multi-agent AI platform.

You will define the architecture from first principles, working directly with the client's senior engineering

leadership.

The platform has two layers: an advisory layer that turns enterprise signals into root-cause analysis,

backlog mapping, and next-best-action recommendations; and an execution layer that models

requirements, self-validates solutions pre-deployment, and orchestrates readiness and go-live.

The defining constraint is governed autonomy — the client operates in a regulated trust-and-

compliance domain, so every autonomous decision must be auditable, factually grounded, and safe.

Key Responsibilities:

• Own end-to-end architecture: ingestion, knowledge/retrieval, agent reasoning, tool access, and governance.
• Design the knowledge lake across telemetry, voice-of-customer signals, CRM, backlog systems, documents, and repositories.
• Architect the MCP gateway, registry, servers, and connectors for governed tool and data access.

Multi-Agent System Design

• Design the multi-agent reasoning core: planner/re-planner, reasoning loops, memory, and agent collaboration.
• Architect solutions utilizing LLMs, RAG, knowledge systems, vector databases, and semantic search.

Governance & Evaluation

• Build governance as a first-class layer: audit trails, grounding, guardrails, human-in-the-loop, and graceful degradation.
• Establish AI governance, security, explainability, observability, and Responsible AI standards.
• Define the evaluation harness, quality metrics, and measurable baselines.

Delivery & Technical Direction

• Drive an end-to-end pilot on a real use case before scaling.
• Set technical direction for the offshore delivery pod.
• Create architecture blueprints, reference implementations, and reusable frameworks; guide engineering teams through implementation and technology decisions.

Required Skills & Experience

• 10+ years in software engineering/architecture, including 3+ years hands-on with agentic or multi-agent LLM systems in production.
• Proven greenfield architecture ownership on a complex, ambiguous system.
• Deep multi-agent orchestration expertise: planning, reasoning loops, tool use, memory, and agent collaboration.
• Strong RAG and knowledge architecture: embedding strategy, hybrid/semantic retrieval, vector stores, and grounding.
• Hands-on experience with agentic frameworks (LangGraph, LangChain, or equivalent) and MCP or comparable tool-access layers.
• Demonstrated AI safety and evaluation work: hallucination mitigation, guardrails, and evaluation harnesses.
• Strong Python and/or Java development experience, with API-first and microservices-based architecture background.
• Exceptional client-facing communication — able to hold your own with a senior technical executive.

Education

• Bachelor’s or master’s degree in Computer Science, Engineering, Artificial Intelligence, or a related field