Technical Architect

Tech Architect Systems Integration & AI Platforms

Looking for someone who gets system integrations at the core: not just the happy path, but where things break, why they break, and how to design so they don't. Someone who has spent years making systems talk to each other across AI, data, and cloud boundaries and has the experience to show for it.

The role covers the full journey, from a one-liner from the business to technical direction, architecture, building the platform, and keeping it running in production. The person we are looking for is who the team comes to when the integration is hard and the answer isn't obvious.

• Experience: 8+ Years
• Work Location: Hyderabad
• Mode of work: Hybrid

What You Will Do

• Start from the business ask. Usually a one-liner. Turn that into a technical direction, get senior stakeholders aligned, and give the engineering team something they can actually build from.
• Own the integration architecture. API contracts, data flow, failure modes, retry behaviour, observability: define these upfront and own them through to production.
• Build platforms other teams build on. Developer-facing API platforms: access tiers, OKTA / OAuth / long-lived credential auth, gateway config, quota management. Needs to be solid, not just working.
• Wire the AI stack into real systems. FastAPI orchestration connecting Bedrock, vector search, document pipelines, analytics, with proper contracts, runbooks, and operational thinking, not just a working demo.
• Build for when things go wrong. Event-driven pipelines on AWS (Lambda, Step Functions, EventBridge, Kafka) with dead letter queues, retry logic, structured logging, and distributed tracing. Not added later, designed in from the start.
• Own production. When something breaks, lead the triage, get to root cause, fix it properly, write the runbook so the next person is faster.
• Set the standard. API contracts, integration patterns, reusable assets, the kind other teams adopt because they are better, not because they are told to.
• Guide frontend and infra. Enough React and Terraform/CI/CD awareness to review and guide delivery, not expected to write production frontend code.

What We Are Looking For

• 8+ years in engineering with real time in architecture or technical leadership, not just the title, actual ownership of how systems are designed and delivered.
• Integration is the core skill here, not a side of the role. Years of making AI, data, cloud, and enterprise systems work together reliably, and knowing exactly where they fall apart.
• Python / FastAPI at depth. Can design, build, and review production FastAPI services without gaps being caught by someone else.
• AWS is the cloud. Lambda, EKS, Bedrock, SageMaker, Step Functions, EventBridge, IAM, VPC: right tool for the problem, and aware of the cost implications of the choice.
• Has shipped infrastructure, not just described it. Terraform, Docker, Kubernetes, CI/CD: part of actual deliveries.
• Security and observability are built in. OAuth, SSO, structured logging, distributed tracing, zero-trust: part of how this person builds, not things added at the end.
• Can read and review a React codebase well enough to guide delivery; writing production frontend code is not expected.

AI & GenAI — What You Need to Know

• No data science needed. But needs to understand the AI stack well enough to make integration and architecture decisions that hold: LLM lifecycles, embedding pipelines, RAG, multi-agent patterns, inference latency, and where each piece fits.
• Vector databases are part of the toolkit. Pinecone, FAISS, Weaviate, Qdrant: knows when to use one and which fits the use-case and scale.
• Can connect an AI system into a production platform: ingestion pipeline, enrichment, post-processing, observability, failure modes. Not just a working API call.
• Responsible AI is an architecture concern. Monitoring pipelines, bias detection, explainability: needs to be designed in, not added later. Should have a clear point of view on how.

How You Work

• Comfortable starting from ambiguity, rigorous by the time it ships. The gap between a vague business ask and a well-designed system is where the best work happens.
• Writes things down. API contracts, architecture decisions, runbooks: not because a process says to, but because it saves everyone time later, including themselves.
• Does not wait for someone to define the standard. If there is no consistent pattern across teams, creates one: good enough that others actually adopt it.
• Collaborates to solve. When something cuts across network, infra, product, or business teams, finds the right people and drives it to a resolution.