AI/ ML Architect (GenAI & LLM Systems)

Location: Hyderabad, Telangana (Hybrid)

Experience: 6+ years in software engineering, with 3+ years in AI/ML and demonstrable ownership of production GenAI systems

Employment Type: Full-Time

Notice Period: Immediate to 30 days preferred

About the Role

We are looking for an AI/ML Architect who can own the technical direction of our AI products end to end from framing an ambiguous business problem, to designing the architecture, to shipping it into production and standing behind it.

This is a senior individual-contributor-plus-leadership role. You will be the person the team turns to for design decisions, the person clients turn to for credibility, and the person the business turns to when it needs to know whether an idea is feasible, what it will cost, and how long it will take. You should be comfortable writing code and equally comfortable defending an architecture in front of a customer’s technology leadership.

What You’ll Own

Architecture & Solution Design

• Own end-to-end architecture for AI/ML and Generative AI platforms- data ingestion, retrieval, orchestration, inference, evaluation, and monitoring.
• Translate ambiguous business requirements into reference architectures, technical specifications, and phased delivery plans.
• Make and defend build-vs-buy, model selection, and hosting decisions (proprietary vs. open-weight, managed vs. self-hosted), backed by benchmarks and cost analysis.
• Define patterns and standards the wider engineering team builds against, so solutions are reusable rather than one-off.

Production, Scale & Cost

• Take models from notebook to production with proper CI/CD, versioning, observability, and rollback paths.
• Own the unit economics of AI features: token and inference cost, caching, batching, model routing/tiering, latency and throughput targets.
• Establish monitoring for drift, degradation, and failure modes, and act on what it reports.
• Architect for scale, reliability, and security across AWS, Azure, or GCP.

Technical Leadership

• Lead and mentor a team of AI/ML and data engineers: design reviews, code reviews, pairing, and raising the technical bar.
• Break down complex initiatives into workstreams, estimate effort credibly, and drive delivery to committed timelines.
• Partner with product, data, and business stakeholders to shape roadmap and to say no to work that won’t deliver value.
• Act as the senior technical voice in client conversations, solutioning workshops, and pre-sales discussions.
• Build the team’s capability - hiring, onboarding, internal knowledge sharing, and evaluating emerging tooling before it becomes a fire drill.

What We’re Looking For

Must-Have

• 6+ years in software/data engineering, with 3+ years building and shipping AI/ML systems that real users depend on.
• Proven ownership of at least one GenAI/LLM application in production - not a prototype, not a POC that stalled.
• Deep, hands-on Python , with production-grade engineering habits (testing, packaging, code quality, version control).
• Strong command of LLMs, embeddings, vector databases, prompt engineering, RAG, and fine-tuning , including the failure modes of each.
• Solid grounding in classical ML, Deep Learning, and NLP
  • you should know when a transformer is overkill.
• Hands-on experience deploying and operating AI workloads on AWS, Azure, or GCP .
• Experience with modern AI tooling - e.g. LangChain/LlamaIndex, PyTorch/TensorFlow, Hugging Face, FastAPI, Docker, Kubernetes, and vector stores such as Pinecone, Weaviate, FAISS, pgvector, or Milvus.
• Demonstrated experience leading or mentoring engineers and owning technical decisions beyond your own code.
• Clear written and verbal communication - able to explain a trade-off to an engineer and to a CFO in the same afternoon.

Good to Have

• MLOps/LLMOps depth: MLflow, Kubeflow, Airflow, LangSmith, Weights & Biases, or similar.
• Experience with model evaluation and observability tooling for LLM systems.
• Exposure to distributed training, model optimisation, quantisation, or inference serving (vLLM, TensorRT, Triton).
• Domain experience in a regulated or data-sensitive industry (BFSI, healthcare, legal).
• Contributions to open source, research publications, patents, or conference talks.
• Cloud or AI certifications (AWS ML Specialty, Azure AI Engineer, GCP Professional ML Engineer).