Balvinder Singh · Blockchain · AI · Architecture
I started in .NET backends and moved through full stack, mobile, and Android payment terminals. Today, as Head of Technology at SeaPos Payment Services, I build across device-agnostic POS, transaction platforms, permissioned blockchain, and AI-assisted commerce.
01 / Highlight
I designed and built a smart-contract escrow and conditional-payments platform on a permissioned Hyperledger Besu (QBFT) network, with banks as custodians and contracts as the decision layer.
Upgradeable Solidity contracts on UUPS proxies, ERC-7201 storage, and OpenZeppelin AccessControl. Invariant and fuzz testing proves that payouts can never exceed secured funds and that valid settlements never get stuck.
An event indexer, keeper, and Custody Gateway that execute bank instructions (fixed-deposit liens, digital rupee) idempotently, and only after re-verifying them on chain.
Off-chain encrypted evidence that can be erased by deleting its key, KMS-managed custodial keys, reconciliation, observability, and a 4-validator fault-tolerant network.
02 / What I bring
Different terminals expose different APIs, device capabilities, and lifecycle constraints. I put those differences behind clear adapters and shared contracts, connecting the application on the device with the services, security controls, and operational visibility behind each transaction.
A common application layer over vendor SDK adapters, with shared payment flows and device-specific capabilities kept explicit.
EMV L2ISO 8583Contactless02 / CORE & ROUTINGTransaction orchestration, status and recovery, secure device connectivity, merchant services, and reporting.
Double-entryHSMReconciliation03 / NEXT & AGENTICPractical experiments in agent permissions, explainable operations, and trustworthy commerce workflows.
ConsentMandatesSettlement03 / Conceptual architecture
Vendor SDKs differ in card-reader APIs, printer behavior, key operations, system services, and lifecycle events. A deliberate adapter boundary absorbs those differences while the shared application owns transaction flow, UI behavior, observability, and product rules.
04 / Working exhibits
Deterministic simulator that shows what Terminal, Middleware, Host, and Merchant know when a transaction takes an unexpected path.
A user mandate meets a proposed transaction. The engine walks every guardrail and produces APPROVE / DENY / STEP-UP with a full audit trail.
Compare payment rails an autonomous agent might use — settlement window, cost, reversibility, identity binding, agent-readiness.
Source-grounded Q&A over a published sample payment architecture. Planned — publishing an AI-backed exhibit requires guardrails first.
Paste an ISO 8583 message, get the MTI, bitmaps, and every data element decoded — BCD or ASCII, TPDU, and EMV field 55. Deterministic, offline.
Paste an EMV hex buffer, get a parsed TLV tree with tag dictionary lookups. Deterministic, offline, safe for production use.
05 / Field notes
06 / Selected work
Founder · Practice lead
Infrastructure engineering and cloud operations partner for teams that would rather ship features than debug their VPC. Consulting, remediation, and on-call.
Open infrabix.com ↗Product · Engineering lead
Project and delivery-tracking product for engineering teams. Tenant-scoped, backup-first, auditable — designed for teams that need real trails, not aesthetic dashboards.
Open gridbix.com ↗Founder · Engineering lead
Government-exam typing practice portal for Indian aspirants. High-throughput, exam-realistic keyboard behaviour, and progress tracking that survives across devices.
Open creativetyping.com ↗Engineering
Small-team cricket scoring, roster, and match-history product. Built for local clubs that want a serious record of their play without spreadsheet gymnastics.
Coming soon07 / Path
Head of Technology
SeaPos Payment Services Pvt. Ltd.
Engineering Manager
Ingenico India International
Technical Lead
Ingenico India International
Lead Engineering
Commdel Consulting Services
Sr. Associate Engineer Lead
Commdel Consulting Services