DEV.co
AI Frameworks · GURPREETKAURJETHRA

END-TO-END-GENERATIVE-AI-PROJECTS

A curated collection of 18+ end-to-end generative AI projects demonstrating LLM applications including RAG systems, chatbots, document processing, and fine-tuning. The repository serves as a reference implementation showcase rather than a production framework, with projects spanning Google Gemini, LLaMA, Mistral, and OpenAI models deployed via Streamlit and FastAPI.

Source: GitHub — github.com/GURPREETKAURJETHRA/END-TO-END-GENERATIVE-AI-PROJECTS
601
GitHub stars
171
Forks
Unknown
Primary language
MIT
License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

FieldValue
RepositoryGURPREETKAURJETHRA/END-TO-END-GENERATIVE-AI-PROJECTS
OwnerGURPREETKAURJETHRA
Primary languageUnknown
LicenseMIT — OSI-approved
Stars601
Forks171
Open issues1
Latest releaseUnknown
Last updated2025-01-24
Sourcehttps://github.com/GURPREETKAURJETHRA/END-TO-END-GENERATIVE-AI-PROJECTS

What END-TO-END-GENERATIVE-AI-PROJECTS is

Repository aggregates working implementations across LLM inference, retrieval-augmented generation (RAG), fine-tuning (LoRA/QLoRA), and multi-modal processing. Core stack includes LangChain, LLaMA Index, Haystack, and vector databases (FAISS, Cassandra/Astra DB) with deployment patterns via Streamlit, FastAPI, and Chainlit. Covers both cloud-hosted (Gemini, OpenAI) and open-source model inference (LLaMA2, Mistral 7B, Gemma 7B).

Quickstart

Get the END-TO-END-GENERATIVE-AI-PROJECTS source

Clone the repository and explore it locally.

terminalbash
git clone https://github.com/GURPREETKAURJETHRA/END-TO-END-GENERATIVE-AI-PROJECTS.gitcd END-TO-END-GENERATIVE-AI-PROJECTS# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Learning LLM application architecture

Reference implementations for building document Q&A systems, RAG pipelines, and multi-turn chatbots. Each project isolates a specific pattern (PDF ingestion, vector search, SQL generation, fine-tuning) suitable for engineers studying LLM application design.

Rapid prototyping of conversational AI features

Projects like Multi-PDFs ChatApp and Medical ChatBot provide working code templates that can accelerate MVP development. Streamlit-based UIs enable quick iteration without building custom frontends.

Exploring open-source vs. proprietary model trade-offs

Parallel implementations using LLaMA2, Mistral, Gemma, and Google Gemini Pro allow direct comparison of inference quality, latency, and operational cost for model selection decisions.

Implementation considerations

  • Each project is independently authored and may use conflicting library versions. No lockfile (requirements.txt/poetry.lock) provided in README; dependency management requires per-project investigation.
  • Projects assume specific hardware (Medical ChatBot requires 16GB RAM minimum). No guidance on scaling inference or batching for concurrent requests beyond single-user Streamlit deployment.
  • Fine-tuning projects (LoRA/QLoRA on LLaMA2) omit critical details: dataset preparation, hyperparameter selection, evaluation metrics, and inference latency post-tuning.
  • Vector database choices vary (FAISS for local, Cassandra/Astra DB for managed). No architecture guide for selecting storage based on scale, latency, or cost constraints.
  • API key management shown in README examples uses environment variables. No secrets rotation, credential lifecycle, or audit trail documentation.

When to avoid it — and what to weigh

  • Production-grade system requiring SLAs — Repository is a collection of examples, not a maintained framework. No versioning, release management, or compatibility guarantees across projects. Unfit for systems requiring uptime commitments or long-term API stability.
  • Sensitive data handling (PII, medical, financial) — Projects lack explicit data governance, encryption, audit logging, or compliance documentation. Medical ChatBot and Resume ATS projects process sensitive information without stated privacy controls—unsuitable for HIPAA, GDPR, or regulated workflows.
  • Integrated deployment across heterogeneous teams — Each project uses different deployment methods (Streamlit, FastAPI, HuggingFace Spaces, Ollama). No unified CI/CD, observability, or operational runbooks. High friction for teams managing multiple projects in production.
  • High-volume, cost-sensitive inference — Projects default to calling cloud APIs (Gemini Pro, OpenAI GPT-3.5) without cost optimization or caching strategies. No benchmarks on token usage, latency, or fallback logic for rate-limited or expensive model calls.

License & commercial use

MIT License. Permits commercial use, modification, and distribution with attribution and no warranty. No patent or trademark clauses. License is clear and OSI-compliant.

MIT license permits commercial use of the repository code itself. However, projects integrate third-party models and APIs (Google Gemini, OpenAI GPT-3.5, LLaMA2, Mistral) with their own commercial terms. Verify each model's acceptable use policy and licensing. Example: LLaMA2 has commercial restrictions on derivatives; some fine-tuning examples may violate those terms depending on intended use.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

SignalAssessment
MaintenanceModerate
DocumentationLimited
License clarityClear
Deployment complexityModerate
DEV.co fitGood
Assessment confidenceMedium
Security considerations

Projects handle sensitive data (medical records, resumes, financial info, SQL queries) without explicit security patterns. No visible handling of input validation, prompt injection mitigation, rate limiting, or access control. API keys embedded in example code or assumed to be in environment—no secure key rotation shown. Vector databases and inference endpoints may be publicly accessible if default configurations are used. Model outputs are not sanitized or validated before presentation to users.

Alternatives to consider

LangChain Cookbook / LangSmith documentation

Official, maintained examples with stronger documentation and production patterns. Covers similar use cases (RAG, agents, fine-tuning) with clearer architectural guidance.

Hugging Face course and tutorials

Comprehensive, versioned learning content for fine-tuning and inference. Includes benchmarks, hardware requirements, and best practices. Better for skill-building than example code.

Ray Serve / BentoML templates

Production-grade deployment frameworks with built-in scaling, monitoring, and A/B testing. Suitable if you need to operationalize a project beyond Streamlit.

Software development agency

Build on END-TO-END-GENERATIVE-AI-PROJECTS with DEV.co software developers

Explore these reference projects to understand RAG, fine-tuning, and deployment patterns. Use as learning material and architectural starting point—then adapt for production with proper error handling, observability, and security controls.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

Related on DEV.co

Explore the category and the services that help you build with it.

END-TO-END-GENERATIVE-AI-PROJECTS FAQ

Can I use these projects as a starting point for production?
Yes for rapid prototyping; no for SLA-bound systems. Expect to refactor deployment, add observability, implement error handling, and test with your own data. Use as architectural reference, not copy-paste template.
Which model should I pick for my use case?
No guidance provided in the repository. Start with the open-source options (LLaMA2, Mistral 7B) for cost/control; fall back to cloud APIs (Gemini, OpenAI) if you need superior quality or reasoning. Run benchmarks on your data.
How do I fine-tune a model for my domain?
Repository includes examples using LoRA/QLoRA on LLaMA2, but omits dataset preparation, evaluation, and hyperparameter tuning. Refer to LLaMA2 fine-tuning guides and Hugging Face documentation separately.
What are the costs of running these projects?
Unknown. Projects using cloud APIs (Gemini, OpenAI) do not estimate token usage or monthly spend. Open-source models (run locally or on Ollama) have compute costs but no benchmarks provided.

Software development & web development with DEV.co

From first prototype to production, DEV.co delivers software development services around tools like END-TO-END-GENERATIVE-AI-PROJECTS. Our software development agency staffs experienced software developers and web developers for custom software development, web development, integrations, and ongoing support across ai frameworks and beyond.

Ready to build your LLM application?

Explore these reference projects to understand RAG, fine-tuning, and deployment patterns. Use as learning material and architectural starting point—then adapt for production with proper error handling, observability, and security controls.