Role description: This role is with one of our portfolio companies.

About the Role

We are seeking a high-conviction AI Architect to define and execute our end-to-end AI strategy, architecture, and platform infrastructure. This role sits at the intersection of deep technical craftsmanship and strategic organizational ownership—combining hands-on architecture of autonomous agentic systems with the leadership required to scale enterprise-grade AI infrastructure.

In this position, you will own the company-wide AI vision, leading the shift from basic LLM integration to production-grade autonomous agent networks, self-correcting cognitive loops, tool-augmented execution models, and scalable platform infrastructure. You will work directly with executive leadership, product, engineering, and security teams to translate ambitious business goals into resilient, deterministic, and high-performance AI solutions.

Key Responsibilities

Agentic AI & Cognitive Architecture

- Autonomous Agent Workflows: Architect, orchestrate, and deploy multi-agent networks (hierarchical, swarm, supervisor-worker patterns) capable of multi-step reasoning, energetic planning, and tool execution.
- Reasoning & Tool Execution: Design robust ReAct, ReFlect, and Plan-and-Execute cognitive loops. Build standard interfaces (including Model Context Protocol / MCP) and sandboxed environments for secure interaction with APIs, databases, code interpreters, and enterprise SaaS.
- Context & Memory Systems: Engineer hybrid state and memory management architectures integrating episodic/procedural memory, short-term context compression, and advanced graph-augmented RAG setups.

AI Infrastructure, Platform & Reliability

- Enterprise Infrastructure: Own, scale, and optimize cloud-native AI platform infrastructure across AWS, GCP, or Azure for low-latency,



high-availability LLM and multimodal inference.
- Model Optimization & Fine-Tuning: Drive model fine-tuning, quantization, distillation, and runtime acceleration using frameworks such as TensorRT, ONNX, vLLM, and open-source model ecosystems (Llama, Mistral, Qwen).
- Observability, Safety & Guardrails: Establish comprehensive AI reliability engineering practices—implementing deterministic guardrails, evaluation pipelines (Ragas, TruLens, DeepEval), telemetry/tracing (LangSmith, Phoenix), and robust defenses against prompt injection and data leakage.

Key Requirements

- Core Experience: 8+ years overall in distributed backend, platform, or cloud infrastructure engineering, with 2+ years of hands-on experience designing, building, and scaling production platforms and agentic workflows.
- Agentic Tech Stack & Frameworks: Deep proficiency with LangGraph, AutoGen, CrewAI, LlamaIndex, or custom state-machine agent orchestrators.
- Model Optimization & Runtime: Direct experience with model acceleration and serving frameworks (TensorRT, ONNX, vLLM, Ollama) along with quantization, fine-tuning (LoRA/PEFT), and distillation techniques.
- Data & Retrieval Architecture: Proven experience with vector databases (Pinecone, Qdrant, Milvus, pgvector) and Knowledge Graphs for complex retrieval setups.
- Cloud & Distributed Systems: Hands-on experience with cloud-native distributed architecture, asynchronous event-driven microservices (Kafka, RabbitMQ), streaming protocols (WebSockets, SSE), and container orchestration (Kubernetes, Docker sandboxes).
- Leadership & Influence: Outstanding stakeholder management skills with a track record of driving cross-functional technical consensus and articulating AI strategy to C-suite partners.
- Education: B.Tech / M.Tech in Computer Science, Artificial Intelligence, Machine Learning, or equivalent practical experience.

📌 Artificial Intelligence Specialist (Bengaluru)
🏢 Elevation Capital
📍 Bengaluru