Head of Data & Analytics @ Onramp
Hesham Meneisi
Hands-on data leader with 8+ years building data platforms for payments, lending, and e-commerce. Usually hired into chaos: near-daily failures, books that don't reconcile. I rebuild the org so agents handle the routine work and the pipelines still reconcile.
Bangkok, Thailand · led teams of up to 7 engineers
// Impact
// Stack
// Experience
Onramp
- Restructured the data function into an AI-first operating model: 3 engineers running 10-15 parallel agents daily, winning executive buy-in by demonstrating critical dashboards built and validated in hours instead of days.
- Held agent-assisted pipelines to audit-grade accuracy: €10M+ monthly volume reconciled across internal data, acquirers, banks, and wallets down to rounding errors, a total discrepancy of under €5.
- Replaced 188 static dashboards with 30 governed self-serve dashboards and an org-wide Claude SQL skill, adopted across operations, finance, compliance, and support, cutting data-team requests from 20+/month to 0-2.
- Built a lakehouse on Azure (300+ governed assets) that retired the fragmented legacy stack and cut infrastructure cost 43%.
- Designed the platform so agents operate it directly: everything as code, fixed repo patterns, and an ADR for every architecture decision. Agents open reviewable cross-repo PRs.
- Automated bi-monthly statement generation and reconciliation (replacing a 2-3 day manual cycle) and standardized financial close across the org.
Honest
- Built a self-service data mesh unifying siloed data into a single cataloged platform.
- Reduced monthly production incidents from 60+ to under 10 via a reliability overhaul and runbook standardization.
- Halved cash-volume loss by reworking the Probability-of-Default credit-approval model: dropping low-signal legacy features, replacing them, and adding a new data source.
- Set up data governance (contracts, SLAs, ownership) and standardized feature definitions across teams.
Omise
- Shipped a third-party clearing/settlement integration in under a month against a multi-month estimate, with a 3-engineer team.
- Rebuilt the payment pipelines: failures fell from near-daily to a few a month and scans dropped from terabytes to gigabytes.
- Deployed Superset dashboards and Spark on Kubernetes for delta-table processing and CDC ingestion.
- Mentored mid-level engineers to senior while the team kept shipping.
CREA
- Optimized a high-volume e-commerce data-sync system, cutting warehouse query time from 2 minutes to 1.5 seconds.
Senior Data & Software Engineer, Devcurate (2019-21, Bangkok), building products end to end · Software Engineer, Leastra (2018-19, remote) · AI & Robotics R&D, Alpha Inference (2020, Geneva) · Algorithms Intern, Valeo (2017, Cairo), cut a sensor-tracking bottleneck by 88%.
// Currently Building
One product a month
I've shipped a new indie product every month since January 2026, across mobile, web, and AI tooling.
Daily AI agents
I keep a few agents running each day for coding, research, and analysis, and build systems that are easy for them to work in.
LLM-native codebases & teams
Shaping repos, docs, and team process around AI agents as everyday contributors rather than bolt-ons.
// Education
Alexandria University
Udacity
IELTS Band 8 (English) · Native Arabic