About the Role

We are looking for an experienced AI Engineer who specializes in building agents and agentic systems—from task-orchestration agents to workflow automation agents, retrieval-augmented agents, research/coding agents, multimodal agents, and domain-specific autonomous agents.

This is a full-stack AI engineering role, ideal for someone who loves shipping: rapid MVPs → stable production, high ownership, and fast problem-solving. Candidates must have built and deployed at least two AI agents in production in the past 12 months and be comfortable operating in high-velocity environments.

Applications are only considered if submitted through our official platform

What You’ll Do1. Build & Deploy AI Agents

Design, build, and ship agentic workflows across multiple domains (research agents, coding assistants, conversational agents (voice, texts, etc), reasoning agents, scheduling agents, analytics agents, workflow automation bots, etc.).

Own the end-to-end lifecycle: data ingestion → reasoning → action taking → evaluation → monitoring.

Build multi-step agents capable of autonomous planning, context tracking, memory, tool use, and API orchestration.

2. Agent Architecture & Infrastructure

Architect systems using modern agent stacks (LangChain, LlamaIndex, OpenAI Assistants, Model Context Protocol (MCP), custom orchestration).

Build robust retrieval pipelines (RAG), vector embeddings, caching layers, and knowledge-grounding systems.

Integrate agents with external tools and systems (APIs, SaaS apps, CRMs, internal services, databases, messaging platforms).

3. Productionization

Deploy agents as microservices with proper observability, evals, guardrails, fallbacks, and monitoring.

Optimize inference cost, latency, accuracy, and task-completion rates.

Run systematic evaluations: function calling accuracy, groundedness, hallucinations, long-context stability.

4. Collaboration & Product Work

Work closely with product managers, domain experts, and engineers to translate business workflows into agent behaviors.

Create reusable frameworks and libraries to accelerate subsequent agent builds.

Document and evangelize agent best practices internally.

Applications are only considered if submitted through our official platform

What You Bring

Required 4–7 years of hands-on experience in AI/ML engineering.

Successful deployment of at least two production AI agents in the past 12 months (not prototypes).

Expertise in:

LLMs: OpenAI, Anthropic, Gemini, Llama, DeepSeek

Agent frameworks: LangChain, OpenAI Assistants, custom orchestration, state machines

Retrieval (RAG), vector DBs (Pinecone, Weaviate, Chroma, PGVector)

API integration & tool-use architectures

Python/Node for server-side agent logic

Microservice deployments (Docker, Kubernetes, CI/CD)

Strong debugging skills across distributed systems, prompt engineering, inference optimization, and agent reasoning traces.

Comfortable building MVPs in days and scaling them to stable production within weeks/months.

Nice to Have

Experience building MCP servers or integrating with MCP tools.

Experience with structured function-calling workflows (JSON schema, tool plans, agent graphs).

Background in building internal agent frameworks or automation engines.

Experience designing evaluation frameworks for agents (task completion metrics, scenario tests).

Familiarity with workflow engines (Temporal, Airflow, Prefect).

Success Looks Like

In your first 3–6 months, you will:

Build and deploy multiple agents that solve real business workflows.

Improve accuracy, response quality, and reliability of existing agents.

Establish a reusable internal agent framework to increase build velocity.

Contribute significantly to cost, latency, and performance improvements.

Become a core owner of agentic architecture and experimentation.

Please Note: All official communications will be sent from our parent company, Vikara (vikara.ai), and candidates are required to create an account on the Aithors platform (powered by Vikara) as part of our hiring process.