ADEBUNMI OLAMIDE
Lagos
Summary
Dynamic Senior AI Full-Stack Engineer with a proven track record at Turing, specializing in Python and GCP. Expertly architected a distributed LLM evaluation platform, enhancing task completion reliability by 18%. Adept at leading cross-functional teams and optimizing complex systems, driving innovation and efficiency in AI solutions.
Overview
7
7
years of professional experience
1
1
Certification
Work History
Senior AI Full-Stack Engineer
Turing
07.2025 - Current
- Architected and deployed a distributed LLM evaluation platform on GCP, orchestrating 200+ concurrent VM workloads via GCP Batch to benchmark frontier AI agents against complex desktop automation tasks at scale.
- Owned the backend evaluation stack end-to-end FastAPI services, PostgreSQL data layer, and Dockerized execution containers serving 10K+ daily evaluation runs with 99.9% task completion reliability.
- Diagnosed and eliminated a systemic scoring contamination bug spanning 4 microservices, recovering ~18% of previously misclassified benchmark results and restoring leaderboard integrity.
- Designed a real-time streaming architecture Pub/Sub listeners, Redis queues, and WebSocket fan-out to surface live agent steps and scoring events to annotators during refinement, cutting evaluation feedback loops from minutes to under one second.
- Led a full architectural migration from REST polling to GraphQL subscriptions, reducing API payload size by ~40% and enabling real-time data synchronization across the entire frontend.
- Shipped a fully automated CI/CD pipeline via GitHub Actions and GCP, eliminating all manual deployment steps and cutting release cycle time by 60%.
- Led cross-functional optimization with the infrastructure team, re-architecting VM provisioning and job scheduling to reduce per-evaluation compute cost by 25% and increase cluster throughput by 3x.
AI Principal Backend Calibrator
Amazon BigCode Project
06.2024 - 07.2025
- Developed the backend and evaluation pipelines using Django/DRF and Celery, supporting both offline and online A/B testing with monitored win rates.
- Managed PostgreSQL data models and integrated ClickHouse for rapid validation of large datasets (class balance, drift, duplicates).
- Implemented a CI quality gate enforcing full test coverage and blocking merges until all pipelines passed; ensured zero unnoticed failures.
- Collaborated with a team of 3 senior and 15 engineers to deliver labeling pipelines with IRR and QA gates maintained on schedule.
- Stabilized production under heavy load by fixing queue starvation, schema drift, and idempotent issues; introduced rate limiting and tracing to sustain p95 latency and error thresholds.
- Conducted large-scale LLM calibration (hard negatives, preference tuning, multilingual evaluation) and merged RLHF data into a unified schema, increasing benchmark win rate by 7-12%.
AI Team Lead Engineer
Microsoft Project
04.2025 - 06.2025
- Built the taxonomy-to-dataset pipeline with validation layers and versioned exports, splitting ~120K items for training and evaluation.
- Designed and maintained L1/L2/L3 metadata taxonomies, integrating automated validators to ensure data consistency across iterations.
- Conducted daily reviews in the labeling tool, enforced IRR and annotation guidelines, and fed edge-case corrections back into taxonomy rules.
- Collaborated with 8 senior engineers and coached 7+ trainers, delivering over 200 curated datasets end to end, fully on schedule.
- Operated LLM calibration pipelines with PostgreSQL and ClickHouse validation, improving online win rate and task completion metrics on targeted evaluation flows.
AI Full-Stack Engineer & LLM Trainer
Apple & ServiceNow projects
06.2024 - 04.2025
- Apple (Python): Developed training and evaluation pipelines for algorithmic data curation and hard negative probing, completing 400+ training cycles with measurable gains in task completion.
- ServiceNow (JavaScript + Function Calling): Built LLM-driven UI flows in Agent Workspace, UI Builder, and Now Experience, ensuring safe parallelism, idempotent, structured error taxonomy, and full audit logging.
- Improved platform performance through GlideQuery tuning, caching strategies, and async worker orchestration.
- Created reusable UI patterns and a lightweight evaluation harness, enhancing development speed and testing efficiency across multiple modules.
Senior Software Engineer
Mindrift
04.2023 - 04.2025
- Rebuilt the labeling tool with a Django/DRF backend, Flask microservices, and PostgreSQL + Redis for scalable evaluations and rubric-driven workflows.
- Integrated LangChain and LangGraph with function calling to enable multi-tool flows (menu lookup, order tracking, refunds, delivery ETA).
- Deployed both local and hosted LLMs with custom guardrails and prompt validation, achieving ~99% guideline conformance on final evaluations.
- Designed training and evaluation pipelines with Celery schedules, offline/live checks, and win-rate dashboards; improved system latency through async I/O, caching, and query optimization.
- Delivered robust APIs and admin UIs supporting idempotent writes, real-time metrics, and alerting, ensuring reliability in production environments.
Full-Stack Developer
Freelance
12.2018 - 12.2024
- Delivered 12+ projects spanning AI applications, e-commerce, and SaaS platforms from concept to deployment.
- Built scalable Next.js + Django applications with MongoDB and PostgreSQL, significantly improving load times and conversion rates.
- Created DeepScreen AI, Proliferate, and KIC Data Privacy tools, driving innovation and adoption across multiple client use cases.
Education
B.Sc. - Computer Science
Oduduwa University
Osun State, NigeriaSkills
- Python
- TypeScript
- JavaScript
- NextJS
- React
- TailwindCSS
- Django
- FastAPI
- Datastar
- Celery
- PostgreSQL
- MySQL
- AWS
- GCP
- ClickHouse
- Redis
- LangChain
- LangGraph
Certification
- HackerRank Software Engineer, 2023
- HackerRank JavaScript Intermediate, 2023
Languages
English
Social Links
xpiralportfolio.vercel.app, linkedin.com/in/xpiral
Timeline
Senior AI Full-Stack Engineer
Turing
07.2025 - Current
AI Team Lead Engineer
Microsoft Project
04.2025 - 06.2025
AI Principal Backend Calibrator
Amazon BigCode Project
06.2024 - 07.2025
AI Full-Stack Engineer & LLM Trainer
Apple & ServiceNow projects
06.2024 - 04.2025
Senior Software Engineer
Mindrift
04.2023 - 04.2025
Full-Stack Developer
Freelance
12.2018 - 12.2024
B.Sc. - Computer Science
Oduduwa University