AI Career Assistant Based on RAG Technology: Practical Analysis of a Telegram Chatbot

This article analyzes a cloud-deployed Telegram AI career assistant chatbot project—AI Career Assistant Chatbot. Combining large language models (LLMs) with Retrieval-Augmented Generation (RAG) technology, the project provides job seekers with precise job recommendations and career consulting services via semantic vector search. It aims to address the pain points of information overload and screening difficulties in the job market, lowering the user threshold through natural dialogue while ensuring the reliability and accuracy of responses.

In the highly competitive job market, job seekers face challenges of information overload and screening difficulties. While traditional recruitment platforms offer a large amount of job information, users need to spend a lot of time browsing and screening. To address this pain point, the open-source project AI Career Assistant Chatbot was developed. This cloud-deployed intelligent job-seeking assistant uses LLM and RAG technologies to provide precise job recommendations and career planning consulting. Its core innovation lies in combining conversational interaction with semantic vector search, allowing users to express their job-seeking needs naturally.

The project is developed using Python 3.8+, with a tech stack including: Telegram Bot API (user interaction interface, multi-device synchronization), LLM (providing intelligent dialogue via OpenAI-compatible API), RAG engine (implementing semantic vector search based on Sentence Transformers), and GitHub Actions (automated continuous deployment). The directory structure is clear, with core modules: chatbot/ (core bot code), JobData/ (job data storage), and .github/workflows/ (automated deployment process). The core method, the RAG engine, implements deep vectorization (converting job descriptions and user queries into high-dimensional vectors), neural search (precisely recalling relevant jobs), and multi-language adaptation (supporting Simplified Chinese and English).

Intelligent Job Recommendations

• Conversational Search: Users describe their needs in natural language (e.g., "Beijing Python backend development, salary 20k+")—the bot understands the intent and returns matching jobs.
• PDF Resume Parsing: Users upload a PDF resume, and the bot automatically analyzes work experience, skills, etc., to identify matching opportunities.

Career Planning and Skill Analysis

Based on real data from the job database, it analyzes core skill requirements for specific positions and provides learning path suggestions.

Interactive Command Support

• /start: Displays welcome message and usage guide
• /job: Switches to job search mode (default)
• /skill: Switches to skill query mode

Reliability Assurance

• Strict Prompt Engineering: Forces the LLM to answer only based on Excel data in the database, prohibiting fabricated information.
• Intelligent Safety Guardrails: Honestly informs users when there are no matching jobs, avoiding misinformation.
• Structured Output: Organizes job name, company, location, deadline, application link, etc., to enhance user experience.

Deployment Operations

• Cloud 24/7 Service: GitHub Actions automatically builds and deploys to cloud servers, ensuring continuous service availability.
• Convenient Data Update: Job data is stored in Jobdata.xlsx—operation staff can update the Excel file and restart the bot to sync data.
• Native Telegram Integration: Users do not need additional apps; multi-device synchronization is achieved via the Telegram client.

Project Value

1. Lowered Usage Threshold: Through the Telegram platform, users can access AI services without installing new apps.
2. Addressing Real Pain Points: Precisely matches job seekers with positions, reducing information asymmetry.
3. Innovative Technology Combination: The RAG+LLM combination ensures response accuracy and natural dialogue experience.
4. Operation-Friendly: Automated deployment and code-free data updates make it easy to maintain and expand.

Outlook

This project is a model of AI technology applied to real scenarios, providing developers with a full-process reference implementation and job seekers with an intelligent job-seeking method. As LLM technology evolves, similar AI assistants will be applied to more fields, and the open-source project also provides a framework for the community to learn and improve.