Contact person: Business Client

Phone:Show

Email:Show

Location: Tripoli, Libya

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"I’m building a chat-style AI agent that can serve as the first line of consultancy across HR, Marketing, and Finance. The assistant needs to answer portfolio-management questions, run financial-risk checks, suggest tax strategies, offer marketing insights, drive strategic analysis, and provide high-level legal guidance—all within one cohesive interface.

Key Responsibilities
RAG System Implementation: Design, implement, and maintain the Retrieval-Augmented Generation (RAG) system, managing the integration between the LLM APIs and the chosen Vector Database. Ensure information retrieval is accurate, relevant, and reliable across all internal documents.

AI Agent Development & Tool Use: Develop and rigorously test AI Agents capable of complex decision-making and task execution. This includes building Function Calling/Tool Use capabilities to enable agents to interface with external APIs (e.g., web search, data analysis services).

Prompt Engineering & Cost Control: Design and manage the comprehensive System Prompts and instruction sets (the "job description" for the AI) to enforce output quality, ethical alignment, and specific consulting formats (e.g., structured reports, bullet points). Proactively optimize prompt design to minimize Token consumption and control variable API costs.

Performance & Scaling: Collaborate on defining and monitoring key performance indicators (KPIs) for the agent system, focusing on response latency, accuracy, and overall system uptime.

Required Technical Skills
Core Languages: Advanced proficiency in Python and relevant libraries (e.g., Pandas, NumPy).

LLM Ecosystem: Direct experience with API usage of leading LLMs (OpenAI, Google Gemini, Anthropic, etc.).

Vector Infrastructure: Hands-on experience with Vector Databases (e.g., Pinecone, Weaviate, ChromaDB) and embedding models.

Frameworks: Proven ability to work with agent development frameworks such as LangChain or LlamaIndex.
A mid-range hourly engagement is fine; I’m expecting a solid, well-documented prototype rather than a polished production rollout. If you have experience shipping multi-domain AI assistants, let’s talk about the best way to bring this to life." (client-provided description)