AI Solution Architect

About the Role :

We are looking for a hands-on AI Solution Architect who can act as the technical face of Antino's AI practice with enterprise clients and also build what is proposed.

You will lead technical discovery and pre-sales discussions, understand business challenges, design AI/GenAI/Agentic AI solutions, estimate effort, explain trade-offs and risks, and work closely with clients, sales, engineering, and data science teams.

This is not a slides-only role. You will build POCs, run demos, write code, define architecture, review production systems, and guide solutions from POC to production.

Key Responsibilities :

1. Pre-Sales & Client Solutioning :

- Lead technical discovery calls, workshops, RFP/RFI responses, proposals, SOWs, estimates, and ROI discussions.

- Translate business problems into practical AI/ML, GenAI, and Agentic AI solutions.

- Present solutions and technical trade-offs to CXOs, architects, security, data, and engineering teams.

- Build rapid POCs and demos to validate and de-risk solutions.

2. Solution Architecture & Delivery :

- Design scalable AI architectures using LLMs, RAG, AI Agents, vector databases, embeddings, prompt/context engineering, and knowledge graphs.

- Select the right approach across classical ML, LLMs, fine-tuning, RAG, agents, or hybrid solutions.

- Build and deploy solutions across AWS, Azure, or GCP.

- Design APIs, data pipelines, model serving, enterprise integrations, and microservices.

- Build production-grade agentic systems including tool calling, multi-agent workflows, memory, human-in-the-loop, and agent hand-offs.

- Establish best practices for security, scalability, performance, observability, evaluation, reliability, and cost optimisation.

3. AI Engineering & Team Enablement :

- Stay hands-on with Python, FastAPI, AI frameworks, LLM platforms, and cloud technologies.

- Guide engineering and data science teams from POC through production.

- Implement AI evaluation, monitoring, guardrails, and LLMOps practices.

- Train internal teams and create reusable reference architectures, accelerators, and playbooks.

- Track emerging AI models, frameworks, protocols, and research and apply relevant innovations to Antino's AI practice.

Must-Have Skills & Experience :

- 5+ years in software engineering, data science, ML engineering, or solution architecture.

- 2+ years of hands-on experience designing and deploying GenAI/LLM solutions in production.

- Strong experience building Agentic AI, including tool-calling agents, multi-agent systems, agentic RAG, memory, and orchestration.

- Strong foundations in statistics, classical ML, deep learning, feature engineering, model evaluation, and time series.

- Strong hands-on Python and API development, preferably FastAPI.

- Experience with LLM platforms such as OpenAI, Azure OpenAI, Anthropic Claude, Gemini, AWS Bedrock, and open-weight models such as Llama, Qwen, Mistral, or DeepSeek.

- Experience with frameworks such as LangGraph, LangChain, LlamaIndex, CrewAI, OpenAI Agents SDK, Claude Agent SDK, or Google ADK.

- Experience with vector/search technologies such as Pinecone, Weaviate, Milvus, Qdrant, pgvector, FAISS, Chroma, Elasticsearch/OpenSearch.

- Strong knowledge of Advanced RAG, GraphRAG, hybrid search, re-ranking, query rewriting, multimodal RAG, and context engineering.

- Experience with MLOps/LLMOps, Docker, Kubernetes, CI/CD, model serving, monitoring, evaluation, and cost optimisation.

- Strong understanding of distributed systems, microservices, databases, APIs, and secure enterprise architecture.

- Strong client-facing and communication skills with the ability to explain complex technical concepts to both CXOs and engineering teams.

Advanced AI Expertise :

- Agent Design : Tool/function calling, ReAct, planning, reflection, memory, state management, HITL, failure recovery.

- Multi-Agent Systems : Supervisor, hierarchical and peer-to-peer patterns, task routing and agent hand-offs.

- Agent Protocols : MCP and A2A.

- RAG : Hybrid search, re-ranking, GraphRAG, knowledge graphs, agentic/multimodal RAG, permission-aware retrieval.

- Model Adaptation : Reasoning, long-context and multimodal models, SLMs, LoRA/QLoRA, distillation, prompting vs RAG vs fine-tuning.

- Inference : vLLM/TGI, quantisation, prompt caching, model routing, latency and token-cost optimisation.

- Evaluation : Ragas, DeepEval, LangSmith, Langfuse, Arize Phoenix, golden datasets, LLM-as-a-Judge, tracing.

- AI Security & Governance : Prompt injection defence, guardrails, PII protection, OWASP Top 10 for LLMs, NIST AI RMF, ISO/IEC 42001, EU AI Act, and India's DPDP Act.

Good to Have :

- Experience in IT services/consulting with global clients across the US, UK, or Middle East.

- Experience with enterprise knowledge platforms, knowledge graphs, or Company Brain-style systems.

- Production experience in Voice AI, Document AI, or Computer Vision.

- Experience with Databricks, Snowflake, BigQuery, or Spark.

- Domain exposure to BFSI, Healthcare, Retail, Logistics, or Manufacturing.

- AWS, Azure, or Google Cloud certifications.

- Open-source contributions, research papers, patents, technical blogs, or conference talks.

Skills

Python, LangChain, LangGraph, Generative AI, Machine Learning, LLM, RAG, Data Science, Agentic AI