- Company Name
- United Software Group Inc
- Job Title
- Python Developer (Agentic AI development)
- Job Description
-
**Job title**
Python Developer (Agentic AI Development)
**Role Summary**
Design, build, and deploy autonomous AI agents that reason, plan, and act within complex environments. Lead development of agent-based architectures using LLMs, tools, memory, and orchestration frameworks to solve real‑world business challenges.
**Expactations**
- Design end‑to‑end agent workflows that integrate LLMs, APIs, databases, and enterprise systems.
- Develop safe, observable, and cost‑efficient agent deployments with human‑in‑the‑loop safeguards.
- Collaborate cross‑functionally with ML, DevOps, and Security teams to produce production‑ready agent systems.
- Stay current on agentic AI research, safety practices, and emerging LLM tooling.
**Key Responsibilities**
- Construct autonomous and semi‑autonomous agents capable of goal planning and decision‑making.
- Build multi‑agent systems that communicate and coordinate actions.
- Integrate LLMs (OpenAI, Anthropic, open‑source) with custom tools, APIs, and vector databases.
- Design and implement agent orchestration frameworks (LangGraph, AutoGen, CrewAI, Semantic Kernel).
- Develop memory subsystems (short‑term, long‑term, vector‑based retrieval) to maintain context.
- Optimize prompts, workflows, and policies for accuracy, reliability, and cost efficiency.
- Implement logging, observability, fallback mechanisms, and human‑in‑the‑loop controls.
- Benchmark agent performance through simulations and real‑world metrics.
- Work with DevOps to containerize, orchestrate, and CI/CD‑deploy agents on cloud platforms.
- Maintain documentation of agent architecture, decisions, and safety protocols.
**Required Skills**
- 8+ years of Python programming; JavaScript/TypeScript a plus.
- 5+ years constructing AI workflows with LangChain, LangGraph, AutoGen, CrewAI, or Semantic Kernel.
- 5+ years designing agent architectures (React, Plan‑and‑Execute, Tool‑Using Agents, Multi‑Agent Systems).
- 5+ years integrating APIs, databases, and external tools into AI workflows.
- 5+ years using cloud platforms (AWS, GCP, Azure) with Docker and Kubernetes.
- 5+ years with vector databases (Pinecone, FAISS, Chroma, Weaviate).
- Strong software engineering fundamentals: version control, testing, CI/CD.
- Preferred: multi‑agent coordination, reinforcement learning or planning algorithms, AI deployment, MLOps, observability tools, AI governance, responsible AI frameworks, NLP/ML/data science background.
**Required Education & Certifications**
- Bachelor’s or Master’s degree in Computer Science, Artificial Intelligence, Machine Learning, or related field.
- Relevant certifications in cloud platforms (AWS Certified Developer, Azure AI Engineer, Google Cloud Professional Data Engineer) or AI/ML (e.g., TensorFlow Developer, PyTorch Practitioner) are advantageous.