Research that keeps actuarial ML honest in time
Mike S. (Mike Swain) — Main Author, Research Programmer, and Scientist at AE / Æ Hive — leads FEATSRV and Feat Mesh 1.0: leakage-free point-in-time feature serving, StochVRP-Mesh organ logistics, open papers, and production-ready APIs.
Mike S.
Main Author · Research Programmer · Scientist · AE Database & Quantum R&D Center · Æ Hive
Designs and ships end-to-end research systems — from temporal theory and synthetic benchmarks to FastAPI surfaces and operator consoles — so actuarial models and organ logistics train and serve on leakage-free, reproducible definitions.
Recommended for you
More from the portfolio
Additional papers, docs, blogging, and open-access archives — latest first.
Latest research & APIs
Papers, public APIs, documentation, and open archives from Mike Swain — Main Author, Research Programmer, and Scientist at Æ Hive. Newest work appears first.
Feat Mesh 1.0 — Point-in-Time Feature Serving Console
Live operator console for online Redis lookups, PIT training previews, leakage audits, and feature importance — authored by Mike S.
FEATSRV: Point-in-Time Correct Feature Serving for Insurance Without Data Leakage
Leakage-free SCD Type 2 as-of joins with Feast, Redis, PySpark, and FastAPI — plus the Nimbus actuarial benchmark.
Feat Mesh 1.0 API documentation
Organ Mesh 1.0 — Advanced Prediction Console for Organ Logistics
StochVRP-Mesh: Stochastic Prediction and NP-Hard Routing for Time-Critical Organ Transplant Logistics
Organ Mesh 1.0 API documentation
Zenodo archive — FEATSRV Point-in-Time report
Zenodo archive — StochVRP-Mesh organ logistics report
Æ Hive Research — blogging, tools, and latest project news
Top research capabilities
Mike Swain advances Æ Hive research through point-in-time stores, Redis online serving, NP-hard organ logistics, leakage science, and open engineering reports.
Feat Mesh 1.0
Live feature serving console — online Redis lookups, PIT training, leakage audits, and importance charts.
FEATSRV
Point-in-time actuarial feature store — Feast, Redis, PySpark, FastAPI with leakage audit and Nimbus packs.
Organ Mesh 1.0
Live organ logistics prediction — dwell, travel, locker diversion, and cold-chain risk synthesis.
StochVRP-Mesh
NP-hard organ logistics — quantile prediction, classical VRPTW, and QAOA pathfinding under ischaemic deadlines.
Open science
Zenodo archives for FEATSRV and StochVRP-Mesh — reproducible engineering reports for peer review.
Research blogging
Figure-led explainers and project news on the Æ Hive Research hub.
Top research products
Recognised among Æ Hive's open research portfolio — advancing hybrid feature serving, organ logistics, and reproducible Zenodo archives.
Feat Mesh 1.0 API
Live feature console — online Redis lookups, PIT training previews, leakage audits, and importance via the Feat Mesh surface.
“GET /online/* — POST /pit-training — leakage audit tables.” — API · Aug 2026
FEATSRV · Point-in-Time
Leakage-free actuarial feature store — Feast, Redis, PySpark, and FastAPI with offline PIT joins and sub-10 ms online lookups.
“Online Redis <10 ms — dual Feast/Dask and PySpark PIT engines.” — NP-hard · Aug 2026
Feat Mesh Docs
Operator and developer documentation for Feat Mesh 1.0 — auth, endpoints, PIT training, and audit workflows.
“Same-origin docs for the live feature console.” — Docs · Aug 2026
Organ Mesh 1.0 API
Live organ logistics console — dwell, travel, and locker diversion predictions with cold-chain risk synthesis.
“POST /predict/* — quantile bars, ischaemic countdown, session log.” — API · Jul 2026
StochVRP-Mesh · Organ Transplant
Quantile dwell/travel/locker prediction fused with classical VRPTW and QAOA pathfinding for ischaemic cold-chain organ delivery.
“Locker ROC-AUC 0.849 — VRPTW + QAOA under ischaemic windows.” — NP-hard · Jul 2026
Organ Mesh Docs
Documentation for Organ Mesh 1.0 prediction routes, cold-chain risk synthesis, and console workflows.
“Guides for operators and integration partners.” — Docs · Jul 2026
Tools + Code
Public APIs, documentation, and open research surfaces from Mike S. — accessible from Æ Hive infrastructure.
Feat Mesh 1.0
Live feature console — online Redis lookups, PIT training, leakage audits, and feature importance.
View project →FEATSRV · Point-in-Time
Leakage-free actuarial feature store — Feast, Redis, PySpark, FastAPI, and Nimbus SCD Type 2 packs.
View project →Feat Mesh Docs
API documentation for Feat Mesh 1.0 endpoints, auth, and operator workflows.
View project →Organ Mesh 1.0
Live prediction console — dwell, travel, and locker diversion with cold-chain risk synthesis.
View project →StochVRP-Mesh Research
NP-hard organ transplant logistics — quantile models, classical VRPTW, and QAOA pathfinding.
View project →Organ Mesh Docs
API documentation for Organ Mesh 1.0 prediction and risk synthesis routes.
View project →GitHub · A-Eq
Open research and platform code from AE Database & Quantum R&D Center.
View project →Get API access
Request production API access for Æ Hive risk and research surfaces.
View project →Risk Management
Financial services risk management — quantum-ready portfolio optimisation workflows.
View project →Mike Swain— Main Author, Research Programmer & Scientist
Mike S. (Mike Swain) leads research programmes at AE Database & Quantum R&D Center / Æ Hive as Main Author, Research Programmer, and Scientist. He architects end-to-end actuarial and logistics experiment pipelines — from temporal feature theory and quantile prediction to FastAPI surfaces, operator consoles, blogging, and open Zenodo archival.
His portfolio spans point-in-time actuarial feature serving (FEATSRV / Feat Mesh 1.0), NP-hard organ transplant logistics (StochVRP-Mesh / Organ Mesh 1.0), leakage science with Nimbus benchmarks, and research hub blogging. Connect via ORCID, LinkedIn, GitHub, or ae@a-eq.com.
Connect on leakage-free feature serving and organ logistics
Papers, Feat Mesh APIs, Organ Mesh consoles, and open-access research programming from Æ Hive.
Stay connected
What's New at Æ Hive — updates from Mike S.'s research programme, Feat Mesh releases, Organ Mesh APIs, and papers.