Backend Engineer

Open to full-time

Bio

Muhammed Y. is a Backend Engineer with several years of experience in building high-performance backend systems and AI-driven applications. He is proficient in Python, FastAPI, and PostgreSQL, and has experience with frameworks like Flask and Django. Muhammed builds APIs, microservices, and data pipelines, applying test-driven development and network programming principles. His expertise extends to database management with PostgreSQL, MySQL, and MongoDB. At Mango Zest Labs, Muhammed architected a Conversational AI engine for virtual call centers, handling high-concurrency sessions with real-time audio streaming. At Fagmaz Hydra Ventures, he improved operational efficiency by 38% through automation and data visibility enhancements. Additionally, he led the backend architecture for a Data-as-a-Service platform at Data Quotient Limited, achieving sub-100ms API response times. Muhammed holds a Bachelor of Science from the University of Osun State. He is well-suited for roles that require building scalable, secure backend systems and AI-driven solutions.

Expertise

PythonFast APIFlaskMongoDBSQLJavaScriptNumPyRelational DatabaseDocument Database

Video Intro

Assessments

General Coding Assessment

(100/100)

Perfect score on H Distance QuestionEasy
Two Algorithm QuestionsMedium
Good score on String Scramble QuestionMedium

Projects

HEX-IAM — Open Source Multi-Tenant Identity & Access Management Platform

• Tech stack: Python, FastAPI, redis, OAUTH 2.0, postgresql, asyncpg, bcrypt, bloom filters • Designed production-grade, multi-tenant IAM platform providing secure authentication, authorization, and identity lifecycle management for SaaS applications • Implemented OAuth/OlDC-compliant authentication flows supporting PKCE, authorization codes, and refresh-token rotation • Built fine-grained authorization with RBAC, FGA, and ABAC-style policies, enforcing strict per-tenant isolation at application and database layers • Designed Redis-backed deterministic token revocation system using Bloom filters for 0(1) revocation checks with constant-time validation, avoiding per-request database calls or introspection endpoints while maintaining strong security guarantees where false positives result in forced re-authentication but never privilege escalation • Enforced strong tenant isolation using PostgreSQL Row-Level Security (RLS), enabling safe multi-tenant operation across shared infrastructure

(Data quotient) Leveler — Multi-Tenant DaaS & Al Platform

• Tech Stack: Python, FastA PI, PostgreSQL, Redis, AWS S3, RAG, Docker • Architected multi-tenant Data-as-a-Service platform aggregating Nigerian and African datasets with market-place for dataset and ML/AI model monetization • Implemented RAG-powered business intelligence with real-time database connectivity for natural-language querying and insight generation • Built configurable, event-driven agentic workflows supporting data enrichment, compliance checks, analytics pipelines, and custom business logic with strict per-tenant isolation

Conversational AI Engine (Mango Zest Labs)

• Tech stack: Python, FastAPI, RabbitMQ, Redis, Open-weight models (faster-whisper, piper-tts, chatterbox-ai, Mistral, Vicuna, Llama 3.2+), RAG, Refrag, postgresql, pgvector • Architected production-grade Conversational AI engine powering virtual call centers and automated customer interaction platforms • Designed layered architecture comprising: SIP ingress layer for call handling, call orchestrator bridging SIP sessions to engine's WebSocket endpoints, STT ingestion for real-time transcription, RAG-based context engine, reasoning/orchestration layer for intent resolution, and TTS voice synthesis • Built scalable backend services coordinating real-time audio streaming, context enrichment, and response generation under high-concurrency call sessions

Demand Forecasting Platform

• Tech Stack: Python, Streamlit, Prophet, XGBoost, Random Forest, Pandas • Built interactive dashboards integrating Prophet, XGBoost, and Random Forest models for time-series forecasting • Delivered real-time demand and inventory forecasts, enabling proactive inventory management and reduced stock-outs by 25%

Domain-specific-search-engine

• Tech Stack: Python, Streamlit, OpenAI API, FAISS • Developed Retrieval-Augmented Generation application enabling natural-language querying over domain-specific datasets • Delivered precise, context-aware answers by combining semantic retrieval with LLM-based reasoning

Experience

Backend Developer
Mango Zest LabsJun 2024 — Sep 2025
- Develop and deploy backend APIs for interacting with Generative AI (GenAI) models
- Build and manage secure authentication layers for AI-driven applications
- Deploy and maintain containerized HashiCorp Vault instances for secure secrets management
- Architected and led development of a production‑grade Conversational AI engine used to power products such as virtual call centers and automated customer interaction platforms
- Designed the system as a layered architecture, comprising: o SIP ingress layer handling inbound and outbound calls o A dedicated call orchestrator that bridges SIP sessions to the engine’s WebSocket endpoints o Speech‑to‑Text (STT) ingestion for real‑time audio transcription o A RAG‑based context engine for domain‑aware knowledge retrieval o A reasoning and orchestration layer for intent resolution, workflow execution, and tool invocation o Text‑to‑Speech (TTS) voice synthesis for natural, low‑latency responses
- Built scalable backend services to coordinate real‑time audio streaming, context enrichment, and response generation while maintaining reliability under high‑concurrency call sessions.
- Developed a stablecoin‑based payment engine enabling low‑fee, near‑instant settlement for customer checkouts.
- Researched and delivered a Web3 proof‑of‑concept for business registration using smart contracts and asset tokenization on the XRP Ledger.
PythonDockerFastAPIOAuth 2.0XRP ledgeropenaiwhispertransformersLLMAgentic AIAI
Python Developer
Fagmaz Hydra VenturesJan 2021 — May 2024
- Created scripts for data processing, task automation, and workflow optimization
- Implemented internal file management systems for secure file sharing
- Improved operational efficiency by ~38% through automation, better data visibility, and workflow optimization.
- Designed and maintained an internal inventory management platform with real‑time analytics, admin endpoints, and automated weekly reporting.
- Performed agricultural and market data analysis to support segmentation, forecasting, and weather‑driven decision making.
PythonPostgreSQLPyQtFASTAPINumpyPandasMatplotlibscikit-learnData AnalyticsflaskRESTApi
Backend Engineer
Data Quotient LimitedJan 2025 — Jan 2026
- Led backend architecture for production-grade Data-as-a-Service (DaaS) and AI platform, achieving sub-100ms API response times under load
- Designed specialized DaaS sub-product from scratch, handling 10,000+ concurrent requests with opGmized query planning and caching
- Implemented NDPR-compliant security controls, authentication flows, and tenant isolation for enterprise clients
- Collaborated with data science and product teams to expose scalable, analytics-ready APIs powering Al-driven insights and ML workflows
PythonFastAPIOAuth 2.0MongoDBPostgreSQLAmazon Document DBAWS S3CloudinaryKotlinKtorAPI integrationData scrapingData Pipelines