Skip to content
View koladefaj's full-sized avatar

Block or report koladefaj

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
koladefaj/README.md

Hi, I’m Kolade 👋

Backend & AI Infrastructure Engineer building production-grade, high-performance systems distributed systems, real-time AI applications, and reliability tooling.

My work focuses on the engineering behind AI products,streaming inference, fault-tolerant task processing, retrieval systems, observability, and backend architectures that continue working under failure.


🧠 What I Do

  • Design fault-tolerant distributed systems with Celery, Redis, and AWS.
  • Build production AI infrastructure including voice AI, retrieval, and LLM orchestration.
  • Optimize latency across real-time inference pipelines.
  • Engineer reliable async backends using FastAPI, gRPC, and PostgreSQL.
  • Build observability into AI systems using OpenTelemetry and Langfuse.
  • Contribute to open-source backend and AI infrastructure projects.

🛠 Tech Stack

Languages
Python (Async) • SQL (PostgreSQL) • Protobuf • Javascript

Backend & Distributed Systems
FastAPI • gRPC • Celery • Redis (Lua) • SQLAlchemy • Docker

AI Systems
LlamaIndex • CrewAI • LiteLLM • Weaviate • pgvector • ONNX Runtime • Langfuse

LLMs OpenAI • Anthropic • Gemini • Ollama

Cloud AWS • GitHub Actions • LocalStack


🚀 Featured Projects

Relier — Zero-Job-Loss Reliability Layer for Celery

Open-source reliability layer implementing the Phoenix Pattern. 100% task delivery under worker failure. pip install relier

Gia — Voice Music Companion Conversational voice AI combining streaming speech, low-latency orchestration, and reflective long-term memory. Features speculative generation and distilled classification to achieve sub-60ms intent routing and ultra-low latency playback, backed by a Weaviate hybrid-retrieval memory consolidation engine.

Phalanx — Real-Time Fraud Orchestration & MLOps Engine
High-throughput microservice ecosystem (FastAPI + gRPC) achieving 66ms transaction SLA and sub-millisecond ML inference. Includes self-healing retraining pipelines, ONNX-optimized XGBoost, agentic AI (Gemma), and AWS Cognito RBAC.

Engram — High-Performance RAG & Document Intelligence
Production-ready Retrieval-Augmented Generation platform with pgvector HNSW indexing, Celery + Redis async pipelines, SHA-256 deduplication, and hardware-accelerated inference (RTX 4060). Features JWT multi-tenancy and zero data-loss guarantees.


🌟 Open Source Contributions

  • Sandhi AI (Distributed Task Queue & System Reliability)
    Took ownership of a major production reliability epic. Delivered 600+ lines of improvements including queue isolation, atomic admission control, circuit breakers, Celery migration, fault-tolerant retries, and overload protection.

  • InclusionAI – AReal (Distributed AI Inference System)
    Refactored RTensor storage endpoints with Pydantic models, migrated legacy Flask blueprints to FastAPI, and introduced reusable schema architecture + enhanced validation.

  • WebTech Network – Autograder System
    Eliminated import-time side effects in Redis driver and introduced constructor-based dependency injection with FastAPI lifespan events.


📫 Let’s Connect


Always open to interesting conversations about distributed systems, AI infrastructure, or backend architecture.
Let’s build something great together! 🚀

Pinned Loading

  1. getrelier/relier getrelier/relier Public

    Zero-job-loss reliability layer for Celery. Phoenix Pattern worker resurrection, atomic idempotency, graceful shutdown, DLQ, OpenTelemetry. pip install relier

    Python 15

  2. gia gia Public

    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 p…

    Python

  3. Phalanx Phalanx Public

    Microsecond-latency ML fraud orchestration engine featuring gRPC microservices, sub-millisecond ONNX inference, and self-healing MLOps workflows.

    Python 1

  4. Engram Engram Public

    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 kn…

    Python

  5. vllm vllm Public

    Forked from vllm-project/vllm

    A high-throughput and memory-efficient inference and serving engine for LLMs

    Python

  6. AReaL AReaL Public

    Forked from areal-project/AReaL

    Lightning-Fast RL for LLM Reasoning and Agents. Made Simple & Flexible.

    Python