Senior AI Engineer - Chatbot & Agentic AI

We are looking for a Senior AI Engineer to join Snoonu's R&D team as a technical anchor for our conversational AI and agentic automation strategy. You will lead the design and delivery of production‑grade chatbot systems powered by AWS and frontier LLMs (Claude‑first), and architect Agentic AI pipelines that encode the company’s SOPs into autonomous, reliable workflows across customer support, order verification, and logistics.

Beyond building, you will set technical standards, guide junior engineers, and act as a key decision‑maker in choosing the next technology directions the team will discover and dominate.

• Own the end‑to‑end architecture of the conversational AI and agentic automation platform — from LLM selection and prompt strategy to cloud infrastructure and observability.
• Define engineering standards, design patterns, and best practices for AI system development; conduct design reviews and enforce quality bars.
• Mentor and guide junior/mid‑level engineers; review code, provide technical feedback, and accelerate the team’s LLM engineering capabilities.
• Partner directly with the R&D Director to evaluate emerging technologies, shape the team’s technical roadmap, and present recommendations with trade‑off analysis.

• Lead the design and delivery of multi‑channel chatbots (web, Whats App, app) using AWS Lex, Bedrock, and API Gateway integrated with Claude or other LLMs.
• Own complex dialogue system challenges: multi‑turn reasoning, context persistence, intent disambiguation, and graceful fallback strategies.
• Integrate chatbots with backend services via secure, scalable RESTful/event‑driven patterns.
• Drive LLM evaluation cycles — benchmark model versions, prompt strategies, and RAG configurations against production quality and cost targets.

• Architect AI systems that encode SOPs as autonomous, multi‑step workflows for customer support, order verification, and logistics operations.
• Select and govern the orchestration approach (Lang Graph, CrewAI, Bedrock Agents, Step Functions) per use case, with clear rationale on reliability, debuggability, and scalability.
• Design robust memory, context management, tool‑use, and guardrail layers to ensure agents behave predictably in adversarial or edge‑case conditions.
• Establish human‑in‑the‑loop checkpoints, confidence thresholds, and escalation paths — ensuring agents augment rather than replace human judgment in critical decisions.

• Design and own the cloud backbone for AI services:
  Lambda, ECS/Fargate, SQS/SNS, Dynamo
  
  DB, S3, Cloud Watch, Bedrock — with focus on scalability, cost, and reliability.
• Build CI/CD pipelines for prompt versioning, model rollout, A/B testing, and automated evaluation before production deployment.
• Define and enforce monitoring standards for drift, latency, cost, and failure rates across all deployed AI systems.

• Lead frontier model evaluation — benchmark Claude, GPT, LLaMA, Mistral, and emerging open‑weight models against specific use cases and constraints.
• Identify and prototype the next high‑leverage AI capability the team should build — bring experiments from idea to validated proof‑of‑concept with clear go/no‑go criteria.
• Produce high‑quality technical documentation: architecture decision records, experimental results, and prompt engineering playbooks for team‑wide use.

• Bachelor’s or Master’s degree in Computer Science, AI, Software Engineering, or a related field.

• 5–8 years of hands‑on software engineering experience, with at least 3 years focused on LLM‑based systems, conversational AI, or agentic architectures.
• Demonstrated track record of owning and shipping production AI systems end‑to‑end — not just models, but the full stack from API to monitoring. Portfolio, Git Hub, or detailed case studies required.
• Prior experience in a senior IC or tech lead role: setting technical direction, conducting design reviews, and mentoring engineers.
• Strong Python and backend development skills (FastAPI/Flask preferred); ability to write clean, production‑grade, maintainable…