Ship better,
from the shipping floor.
Deep, opinionated guides from the senior engineers at Origami — mobile, AI, gaming, auctions, fintech and everything in between.
How to build a RAG application in 2026
The end-to-end guide to retrieval-augmented generation — chunking, embeddings, hybrid search, reranking, evals.
How to choose between React Native and Flutter
A practical framework for picking your cross-platform mobile stack in 2026.
How to hire mobile app developers (2026 guide)
What to screen for, red flags, and how to structure a senior-only engagement.
Class II vs Class III gaming — a plain-language explainer
Server-drawn bingo, RNG, GLI standards and what tribal vs commercial operators need.
How to build an auction platform from scratch
Anti-snipe, soft-close, proxy bidding, escrow, KYC and live-stream — the full auction engine playbook.
LLM observability in production
Tracing, evals, token budgets, cost ceilings and graceful degradation for LLM apps behind real logins.
Next.js 14 App Router in production
Server components, edge runtimes, ISR, and the mistakes teams make on their first App Router site.
Designing agentic LLM workflows
Tool-calling, retries, budgets, eval harnesses, and where agent designs commonly fail.
Programmatic SEO for B2B — a real playbook
How to scale from 40 to 2,000+ keywords without publishing junk.
KYC / AML for consumer fintech apps
How to layer identity, sanctions and risk without wrecking activation.
Postgres vs MongoDB in 2026 — an engineering POV
When to reach for Mongo, and why we still start almost every project on Postgres.
Building a server-authoritative bidding engine
Latency, fairness, anti-snipe, tie-breaking and why the client is never the boss.
Mobile app performance checklist for 2026
Cold-start budgets, 60fps lists, Hermes, JSC, and the metrics that actually predict retention.
ChatGPT SEO — how to become the answer AI cites
llms.txt, ai.txt, JSON-LD, entity SEO, and how to earn citations from ChatGPT, Perplexity and Google AI Overviews.
Prompt caching and LLM cost control
Anthropic prompt caching, OpenAI cached prompts, and the cost-per-request tricks that keep AI features profitable.
Voice agents with OpenAI Realtime
Low-latency turn-taking, barge-in, function-calling and telephony bridges for real voice AI.
Cross-border payouts for creator + fan apps
Stripe Connect vs Wise vs Airwallex — routing payouts to global creators without wrecking margin.
Content moderation with LLMs — what actually works
Where classical ML wins, where LLMs shine, and how to layer them without losing recall on edge cases.
WebRTC live-streaming for auctions
Sub-second latency, phone bidders, auctioneer paddles and the actual protocol choices behind live sales.
React Native vs native — the honest 2026 tradeoff
When to reach for Swift/Kotlin, when RN's ecosystem outperforms native, and how to hedge.
pgvector vs Pinecone — vector search in 2026
Cost, latency, filtering and the reasons most teams should start on pgvector.
Header bidding, explained plainly
Prebid, wrapper tags, SPO, and why publishers reach for header bidding over waterfall.
How server-drawn Class II bingo actually works
Pattern determination, draw servers, RGS, and the compliance layer around tribal Class II gaming.
GLI-19 compliance checklist for RMG operators
What Gaming Laboratories International tests for, and how to build so audits are boring.
Expo + EAS in production
Managed workflow at scale — updates, credentials, submissions, and where you'll eventually eject.
SwiftUI in production — 2026 edition
Widgets, Live Activities, StoreKit 2, and where UIKit still wins.
Jetpack Compose in production
Modern Android UI at scale, performance rules, and Compose-for-multiplatform in real projects.
Shipping B2B SaaS onboarding that converts
Time-to-value, empty states, and the onboarding flows we've measured across dozens of launches.
A pragmatic data warehouse for consumer apps
Segment or Rudderstack → Snowflake / BigQuery, dbt for modeling, and the dashboards you actually need on day one.
Shipping mobile releases weekly without burning out
Trains, phased rollouts, TestFlight, Play internal — how modern mobile shops ship every 7 days.
Designing for tribal casino players
Accessibility, hand economics, session length — the design POV we bring to Class II floors.
Trust & safety for marketplaces
Fraud, identity, disputes, refunds — the guardrails every consumer marketplace eventually builds.
Product analytics stack for mobile apps in 2026
Amplitude, Mixpanel, PostHog, RudderStack — what to pick, and how to avoid analytics debt.
Shipping multi-tenant SaaS on Postgres
RLS, schema-per-tenant, role explosions, and cost management as you grow.
HIPAA-compliant app development, without the theatre
BAAs, encryption at rest and in transit, audit trails, and where the compliance is actually earned.
Structured outputs for LLM apps
JSON mode, function calls, grammars — and how to get 100% valid outputs without brittle prompts.
Fraud patterns in real-money gaming
Bonus abuse, collusion, chip-dumping, mule accounts — and the risk engine that catches them.
Cursor-based pagination for social feeds
Why offsets break at scale, and the cursor patterns that don't.
Cost per inference for LLM features
Modeling the actual unit economics of LLM features — cache hits, routing, and cheap-shot fallbacks.
How to hire AI developers who ship
What to screen for beyond 'has used ChatGPT' — and the take-homes that actually predict output.