New opportunity

Solution Architect

TechDome · Noida, India

About this role

Role Overview: TechDome is seeking a hands-on Solution Architect to lead the end-to-end architecture of a multi-tenant AI platform. The role focuses on building robust infrastructure including the integration bus, semantic layer, agent runtime, workflow engine, and data stores. Key Responsibilities: You will own the target architecture and architecture decision records (ADRs), finalize the durable workflow engine and tenancy model, and design the semantic resolution engine. You will define API contracts, event schemas, and security standards while conducting design reviews and ensuring the platform supports up to 10,000 concurrent runs. You will also manage scalability, tenant isolation, latency, and disaster recovery. Required Qualifications: Candidates must have 15+ years of experience designing and shipping scalable, multi-tenant SaaS or platform products. You need strong knowledge of distributed systems, durable workflows, Kafka, and idempotency. Production experience with LLM-based systems, agent orchestration, LangGraph, RAG, and evaluation is required. You must have enterprise integration experience (CRM, ERP, REST, SOAP, JDBC) and a deep understanding of B2B security architecture, RBAC, and PII handling. Hands-on coding ability is essential. Preferred Qualifications: Experience with Temporal, Apache Camel, Kubernetes operations at scale, and a history of technical talks or open-source contributions are considered assets. Key Focus Areas: Multi-tenancy, enterprise security, distributed systems, durable workflow execution, AI agents, semantic resolution, RAG, and data architecture.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

Distributed SystemsMulti-tenant SaaSKafkaRelational DatabasesIn-memory DatabasesVector DatabasesLLM-based systemsTool callingAgent orchestrationLangGraphRAGEnterprise integrationRESTSOAPJDBCB2B security architectureRBACTemporalApache CamelKubernetes

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Architecting Multi-Tenant AI Infrastructure
The role requires immediate ownership of the platform's core architecture, specifically finalizing the durable workflow engine, tenancy model, and data-store boundaries within the first two weeks. You must ensure the system supports 10,000 concurrent runs while maintaining strict tenant isolation and B2B security standards.
Semantic Resolution and Agent Orchestration
You are responsible for designing the semantic resolution engine that translates plain-language definitions into actionable system tools and access rights. This involves deep integration with LLM-based systems, LangGraph, and RAG frameworks to ensure reliable agent behavior and tool calling.
Enterprise Integration and Data Reliability
You must bridge modern AI capabilities with legacy enterprise environments (CRM, ERP, SOAP, JDBC). This requires designing robust event schemas and API contracts that handle complex data flows while ensuring idempotency, back-pressure management, and disaster recovery.

Likely challenges to think through

Balancing Rapid Delivery with Architectural Rigor
The requirement to finalize core architectural components within two weeks creates a high-pressure environment. Prepare by being ready to articulate your decision-making framework for ADRs and how you prioritize trade-offs between speed and long-term system maintainability.
Managing Complexity in Distributed AI Systems
Integrating durable workflows with non-deterministic LLM outputs poses significant reliability challenges. Prepare by demonstrating how you handle failure states, state management in agentic workflows, and the specific strategies you use to ensure consistency across distributed components.
Strict Security and Compliance in Multi-Tenancy
Handling PII and RBAC within a shared AI platform requires rigorous isolation. Prepare by discussing your experience with B2B security architecture, specifically how you enforce tenant boundaries at the data and execution layers without compromising system performance.

Practice your answers

Technical practice

Designing Multi-Tenant AI Platforms

What to coverExplain your approach to ensuring strict tenant isolation in a shared AI platform. Discuss the trade-offs between logical isolation (e.g., RBAC, schema-level) and physical isolation (e.g., dedicated clusters) when handling sensitive PII and B2B security requirements.

Durable Workflow Execution

What to coverHow do you ensure idempotency and state consistency in long-running AI agent workflows? Discuss the role of durable workflow engines like Temporal or custom saga patterns in managing failures and retries within distributed systems.

Semantic Resolution Engine Design

What to coverDescribe how you would architect a semantic resolution engine that translates natural language into system tool calls and access rights. Focus on the integration between the LLM layer, the supervisor graph, and the underlying enterprise data stores.

Scalability and Back-pressure in Kafka

What to coverWhen processing up to 10,000 concurrent runs, how do you manage back-pressure and throughput in a Kafka-based event architecture? Address how you handle message ordering and consumer group scaling to prevent system bottlenecks.

RAG and Vector Database Strategy

What to coverCompare the performance and architectural trade-offs of different vector database indexing strategies for RAG applications. Discuss how you maintain data freshness and handle retrieval latency in a high-concurrency production environment.

Enterprise Integration Patterns

What to coverHow do you design a robust integration bus to handle diverse legacy protocols like SOAP, JDBC, and REST? Suggest an architectural pattern for normalizing these disparate data sources into a unified schema for AI consumption.

Agent Orchestration and LangGraph

What to coverExplain the architectural challenges of managing state in complex agentic workflows using frameworks like LangGraph. Discuss how you handle state persistence, graph transitions, and error handling across distributed nodes.

Disaster Recovery for AI Infrastructure

What to coverWhat is your strategy for disaster recovery and data retention in a platform that relies on both relational databases and vector stores? Suggest how to balance RTO/RPO requirements with the high cost of re-indexing vector data.

Behavioural practice

Managing Technical Debt vs. Delivery

What to coverDescribe a situation where you had to balance the need for rapid feature delivery with the necessity of maintaining high architectural standards. How did you document these trade-offs in ADRs to ensure long-term maintainability?

Collaborating with External Stakeholders

What to coverProvide an example of a time you had to align your platform's security architecture with the requirements of a pilot customer's security team. How did you negotiate constraints while maintaining the integrity of your multi-tenant model?

Mentoring and Design Reviews

What to coverTell me about a time you conducted a design review that significantly changed the direction of a project. How did you provide constructive feedback to the engineering team while ensuring the platform's scalability and reliability goals were met?

Handling Critical Production Failures

What to coverDescribe a scenario where a distributed system you owned experienced a major failure. What was your process for root cause analysis, and what architectural changes did you implement to prevent recurrence?

Sharpen your knowledge

Topics to prepare

Multi-tenant SaaS Architecture
Review strategies for tenant isolation, data partitioning, and resource governance at scale. Exercise: Draft an Architecture Decision Record (ADR) comparing logical vs. physical isolation for a multi-tenant AI platform supporting 10,000 concurrent runs. This is critical for the role's requirement to own the platform's tenancy model.
Durable Workflow Execution
Study patterns for handling long-running processes, sagas, and idempotency in distributed systems. Exercise: Design a state machine for an agentic workflow that ensures consistency across failures using Kafka and a durable engine. This directly addresses the requirement to finalize the platform's workflow engine.
B2B Security and PII Handling
Examine enterprise security standards, RBAC, and secrets management in the context of AI systems. Exercise: Map out a security architecture for a RAG pipeline that ensures PII is scrubbed before ingestion and access is strictly controlled via RBAC. This is essential for the role's focus on enterprise-grade security.

Skills to revise

Agent Orchestration with LangGraph
Practice building supervisor graphs and managing state transitions between AI agents. Exercise: Implement a simple multi-agent workflow using LangGraph that handles tool calling and error recovery. This is a core technical requirement for the AI platform architecture.
Enterprise Integration Patterns
Refine skills in connecting disparate systems using REST, SOAP, and JDBC. Exercise: Create a schema for an integration bus that normalizes data from a legacy ERP and a modern CRM into a unified format. This is necessary for the platform's integration bus responsibilities.
Semantic Resolution Engine Design
Develop methods for translating natural language into actionable system commands and tool calls. Exercise: Prototype a semantic layer that maps user intent to specific API contracts and access rights. This is a specialized requirement for the platform's semantic resolution engine.

Your preparation checklist

Prepare an Architecture Decision Record (ADR) sample
Draft a sample ADR that demonstrates your process for evaluating trade-offs in a distributed system, specifically focusing on a complex challenge like tenant isolation or durable workflow state management in a multi-tenant SaaS environment.
Develop a technical walkthrough of an AI agent system
Prepare a detailed explanation of a past project involving LLM-based systems or agent orchestration. Focus on how you handled tool calling, RAG pipelines, and the specific frameworks (like LangGraph) used to ensure reliability and idempotency.
Map enterprise integration strategies
Create a summary of your experience integrating legacy enterprise systems (ERP, CRM) using REST, SOAP, and JDBC. Be ready to discuss how you maintain security, handle PII, and manage data consistency across these disparate sources.
Formulate questions on platform scalability
Prepare specific questions regarding the current state of TechDome's platform, such as: 'How are you currently handling back-pressure and failure recovery for the 10,000 concurrent runs target?' or 'What are the primary bottlenecks in your current semantic resolution engine?'
Review B2B security and RBAC patterns
Refine your ability to articulate how you implement multi-tenant security, including secrets management and granular RBAC, ensuring you can explain how these standards are enforced across an AI-driven architecture.

Availability has not yet been verified. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in India. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…