Skip to content

Journal

Field notes

What we learned shipping finance and realtime products — without the agency fluff.

01 · 2026-09-29

WebSocket Fanout Runbook for a Micron Night

Ops runbook style notes for quote fanout when MU prints Wed Sep 30 AMC — what we rehearse with clients.

Read →

02 · 2026-09-29

Top 10 Questions Before Hiring a Dev Shop

The questions that separate shops that ship from shops that invoice.

Read →

03 · 2026-09-29

The Software Is the Cheap Line

2026 quotes for a trading MVP span five figures to six. The spread is scope, and the scope that hurts is not the screens.

Read →

04 · 2026-09-29

OpenAI DevDay (Sep 29): Desk Takeaways for Fintech Builds

What our two-person desk will actually consider shipping after DevDay — and what we will refuse to bolt onto an order path.

Read →

05 · 2026-09-29

7 Red Flags in Agency Contracts

The contract clauses that predict a bad engagement before it starts.

Read →

06 · 2026-09-29

5 Ways to Scope an MVP (and the One That Works)

Every MVP-scoping method, ranked by how often it produces a shipped product.

Read →

07 · 2026-09-28

When We Reach for Skia on Order Books

Desk rule: Skia for dense painted books; FlashList for lists. How we decide inside a fixed-price sprint.

Read →

08 · 2026-09-28

JSI Sockets vs JS WebSockets in Client Sprints

How we choose socket stacks under fixed scope — profiling first, fashion never.

Read →

09 · 2026-09-28

The Data Contract Is the Product

In trading apps, the API contract is the architecture. Define the data first or the UI invents it.

Read →

10 · 2026-09-27

Symbol Sharding RFC We Hand Clients Before Earnings Floods

The one-page RFC we attach to discovery when a client’s watchlist will face AMC floods.

Read →

11 · 2026-09-27

The Scope Firewall: Saying No on Trading-App Builds

"Can we add crypto too?" is how trading apps die. The scope-firewall rule we run on every build.

Read →

12 · 2026-09-27

Feature Flags for Earnings Alert Banners

Banner config for event nights (MU, NKE) — flags, copy review, and no invented numbers.

Read →

13 · 2026-09-27

What Belongs in Discovery Before a Paper-Trading Sprint

Two weeks to name the data contract, the ledger, and the line we will not cross. Then a fixed price means something.

Read →

14 · 2026-09-26

Acceptance Tests for Watchlists Under AMC Bursts

Test plan language we put in work orders so ‘realtime’ means something on Micron night.

Read →

15 · 2026-09-26

The iOS 27 Rebuild: What Agencies Are Charging For Right Now

September 14 forced a rebuild wave. What the work actually involves — and what it should cost.

Read →

16 · 2026-09-26

Our Expo OTA Policy When Markets Move Overnight

Desk policy: OTA for JS/copy; store review for native. Overnight prints do not create exceptions.

Read →

17 · 2026-09-25

Hiring Notes: What We Look for in Trading-App Engineers

We staff sprints for trading products. These are the signals that predict success.

Read →

18 · 2026-09-25

Observability When VIX Is Quiet and Single Names Are Loud

Dashboards we insist on: per-symbol latency, reconnect storms, JS hitches — especially in calm index weeks.

Read →

19 · 2026-09-24

Why Fintech Clients Get Fixed-Price Sprints

Hourly billing rewards slowness. In fintech it also punishes the client's biggest risk: scope drift.

Read →

20 · 2026-09-23

Discovery for Paper Trading Apps: The Questions That Matter

Before a line of code: the eight questions that decide whether a trading-app build succeeds.

Read →

21 · 2026-09-22

Hire Notes: People Who Can Ship Paper-Trading Loops

How we staff sprints that look like Stock Picks — ledger honesty over demo chrome.

Read →

22 · 2026-09-22

Case Study: The Earnings Dashboard Sprint

A three-week fixed-price build — calendar, estimates, implied moves — and what we refused to build.

Read →

23 · 2026-09-20

Why We Only Do Fixed-Price Sprints

Hourly billing rewards slowness. Here's the alternative structure that keeps everyone honest.

Read →

24 · 2026-09-12

What We Learned Building a Paper Trading Platform

Market data contracts, ledger correctness, and why the boring parts are the hard parts.

Read →