koladefaj/JamAIBase
The collaborative spreadsheet for AI. Chain cells into powerful pipelines, experiment with prompts and models, and evaluate LLM responses in real-time. Work together seamlessly to build and iterate on AI applications.
Backend & AI Engineer | FastAPI • Celery • Redis • gRPC | Building Production RAG, MLOps & Distributed Systems at Scale
Repositories
koladefaj/litellm
Python SDK, Proxy Server (AI Gateway) to call 100+ LLM APIs in OpenAI (or native) format, with cost tracking, guardrails, loadbalancing and logging. [Bedrock, Azure, OpenAI, VertexAI, Cohere, Anthropic, Sagemaker, HuggingFace, VLLM, NVIDIA NIM]
koladefaj/vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
koladefaj/AReaL
Lightning-Fast RL for LLM Reasoning and Agents. Made Simple & Flexible.
koladefaj/sglang
SGLang is a high-performance serving framework for large language models and multimodal models.
koladefaj/StoryCanon
A local, privacy-first writing tool that catches continuity errors in fiction and lets the author decide what's canon.
koladefaj/autograder
Flexible, Consistent and Powerful Autograding Tool that grades and generates reports on your students submissions.
koladefaj/embeddedllm
EmbeddedLLM: API server for Embedded Device Deployment. Currently support CUDA/OpenVINO/IpexLLM/DirectML/CPU
koladefaj/koladefaj
koladefaj/gia
Conversational voice AI with streaming STT/TTS, speculative execution, and a distilled routing cascade. Features background memory consolidation using Weaviate hybrid retrieval and mood inference pipelines via Celery.
koladefaj/Phalanx
Microsecond-latency ML fraud orchestration engine featuring gRPC microservices, sub-millisecond ONNX inference, and self-healing MLOps workflows.
koladefaj/Engram
Hardware-accelerated RAG pipeline and document intelligence engine. Features pgvector-native storage, asynchronous OCR workers, and local LLM orchestration via Ollama. Built for high-concurrency knowledge retrieval.
koladefaj/Rapid-MLX
The fastest local AI engine for Apple Silicon. 4.2x faster than Ollama, 0.08s cached TTFT, 100% tool calling. 17 tool parsers, prompt cache, reasoning separation, cloud routing. Drop-in OpenAI replacement. Works with Claude Code, Cursor, Aider.
koladefaj/sandhi_ai
koladefaj/e-pharmacy-backend
A full-featured e-pharmacy backend API built with FastAPI, featuring role-based access control (User/Pharmacist/Admin), prescription verification workflow, Stripe payment integration and asynchronous task processing with FastAPI background task.
koladefaj/Telco-churn-prediction
Machine learning project predicting telecom customer churn using a tuned XGBoost model with custom threshold optimization, EDA, and deployment-ready pipeline.
koladefaj/data-wrangling
Beginner-friendly data wrangling and visualization projects using Python (Pandas, NumPy, Matplotlib)