redfly.ai is built with Angular, hosted on Azure Static Web Apps. It sends email through Twilio SendGrid.
| Technology | Version | First seen |
|---|---|---|
| JavaScript Frameworks & Libraries1 | ||
| 18.2.14 | 9 Nov 2026 | |
| CDN & Hosting2 | ||
| – | 9 Nov 2026 | |
| – | 9 Nov 2026 | |
| Transactional Email1 | ||
| – | 9 Nov 2026 | |
| Fonts & Icon Fonts1 | ||
| – | 9 Nov 2026 | |
Registration from the registry's RDAP record, the certificate the site serves, and its rank in public traffic lists.
The crew hasn’t read this site’s colors yet. We read a palette when we capture a site’s screenshot, busiest sites first. Meanwhile, browse palettes from other sites.
From what the site says about itself, and account IDs it shares with other sites.
redfly-ai.mail.protection.outlook.compdns2.registrar-servers.compdns1.registrar-servers.comv=spf1 include:spf.protection.outlook.com include:sendgrid.net -all_j2cki981t7ln56r64j6wr41zas00t86openai-domain-verification=dv-L19m1u5eL6TCRjoNVE0LVK1Xv=DMARC1; p=quarantine; pct=100# redfly.ai: every page is public and may be indexed by every crawler, including AI crawlers. # Machine-readable summaries: /llms.txt (short) and /llms-full.txt (every article in full). # Article feed: /feed.xml. Sitemap: /sitemap.xml. User-agent: * Allow: / # Search engines User-agent: Googlebot Allow: / User-agent: Bingbot Allow: / User-agent: DuckDuckBot Allow: / User-agent: Applebot Allow: / User-agent: YandexBot Allow: / User-agent: Baiduspider Allow: / # AI assistants and their search crawlers User-agent: GPTBot Allow: / User-agent: ChatGPT-User Allow: / User-agent: OAI-SearchBot Allow: / User-agent: ClaudeBot Allow: / User-agent: Claude-User Allow: / User-agent: Claude-SearchBot Allow: / User-agent: anthropic-ai Allow: / User-agent: PerplexityBot Allow: / User-agent: Perplexity-User Allow: / User-agent: Google-Extended Allow: / User-agent: Applebot-Extended Allow: / User-agent: Amazonbot Allow: / User-agent: Meta-ExternalAgent Allow: / User-agent: Meta-ExternalFetcher Allow: / User-agent: Bytespider Allow: / User-agent: CCBot Allow: / User-agent: cohere-ai Allow: / User-agent: DuckAssistBot Allow: / User-agent: YouBot Allow: / User-agent: MistralAI-User Allow: / User-agent: PetalBot Allow: / User-agent: Diffbot Allow: / Sitemap: https://redfly.ai/sitemap.xml
# redfly.ai > redfly works with warehouse and distribution companies to effectively integrate AI into their existing systems. When an integration needs data from several sites, the warehouse sync add-on provides it: a small service at each site watches the tables the customer chooses in the site's SQL Server, MongoDB or Postgres database and sends changes outward only, compressed and encrypted with a key unique to that site, into one central store in the cloud that holds the current state of every table, every row stamped with its site, current to within a couple of minutes. It queues through outages, resumes per table, throttles itself dynamically to preserve the performance of the site database under load, and needs no change to the software running the site. The customer builds its own reports, dashboards and analytics on top. The second product, the redfly API, serves an application's reads from a memory cache (Redis) kept in sync with the database and passes writes straight to the database, cutting the database line of the cloud bill. Key facts: site sync sources are SQL Server (change tracking read on a short interval), MongoDB and Postgres; the destination is low-cost document storage (MongoDB) holding the current state of each table; delivery is at-least-once, so nothing is lost; one-way, outbound only, nothing reaches in from outside. Our team has synced more than ten billion rows across more than one hundred sites over ordinary public internet links; production grade. The caching product supports SQL Server, MongoDB and PostgreSQL, offers REST and gRPC APIs with Get, Insert, Update, Delete, GetRows and GetTotalRowCount, serves joins and multi-field queries from Redis, and falls back to the database automatically. Deployment: the customer's own cloud account (preferred), fully managed by redfly, or onsite; no dependency on any cloud vendor's proprietary services. Design-partner contract that includes the architecture review, followed by an annual or monthly subscription; no free trial. ## Pages - [Home](https://redfly.ai/): AI for Logistics: redfly works with warehouse and distribution companies to effectively integrate AI into their existing systems. - [How it works](https://redfly.ai/how-it-works): Where the AI sits in a logistics operation and the seven steps redfly takes with you to get one integration live. - [Warehouse sync](https://redfly.ai/logistics/warehouse-sync): An add-on: one central copy of every warehouse's data, current to within a couple of minutes, with reporting load off the onsite servers. Used when an AI system or report needs data from several sites. - [How warehouse sync works](https://redfly.ai/logistics/warehouse-sync/how-it-works): What runs at the site and in the cloud, what a site needs, the data path, behaviour when the link drops, load management, the central store, boundaries, and who runs what. - [Caching](https://redfly.ai/caching): The second redfly product: an API that serves reads from a memory cache kept in sync with the database and passes writes straight to the database; cuts the database line of the cloud bill and makes reads far faster than hand-written data access code. - [How caching works](https://redfly.ai/caching/how-it-works): The redfly API, Local Sync beside the database, change capture per engine, automatic fallback to the database, and what the customer keeps out of their codebase. - [Caching use cases](https://redfly.ai/caching/use-cases): Where the caching product fits: cloud cost reduction, scaling read-heavy applications, legacy modernization, IoT and fleet telemetry, retail peaks, analytics and dashboards. - [Caching pricing](https://redfly.ai/caching/pricing): The design-partner contract for the redfly API and caching, and the deployment options (your own cloud account, fully managed by redfly, or onsite). - [Insights](https://redfly.ai/insights): The AI for Logistics articles: where AI fits in a logistics operation, what the data needs, where it runs, and how to get one integration live. - [Warehouse sync insights](https://redfly.ai/insights/warehouse-sync): Field-tested notes on moving data from remote sites to the cloud: offline support, peak seasons, per-site encryption and what a central copy is for. - [Caching insights](https://redfly.ai/insights/caching): Field-tested notes on serving reads from memory, cutting the database line of the cloud bill, and the engineering behind it. - [About](https://redfly.ai/about): Who builds redfly and the two decades of data sync and data platform work behind it. - [Investors](https://redfly.ai/investors): The AI for Logistics story for investors: why now, the market, what redfly builds, how it makes money, the team, the track record and the Seed ask. - [Contact](https://redfly.ai/contact): How to reach the redfly team. - [Privacy](https://redfly.ai/privacy): Privacy policy. - [Terms](https://redfly.ai/terms): Terms of use. - [License](https://redfly.ai/license): Software license terms. ## Articles - [Seven steps from “where does AI fit” to a live integration](https://redfly.ai/insights/seven-steps-from-where-does-ai-fit-to-a-live-integration): How redfly takes one AI integration from a first conversation to live use across your operation, which of the seven steps we share with you, and what your people do at each one. (AI for Logistics, 2026-10-31) - [The redfly Portal: where alerts, suggestions and dashboards land](https://redfly.ai/insights/the-redfly-portal-where-alerts-suggestions-and-dashboards-land): AI output only helps when someone can see it and act on it. Here is what lands in the redfly Portal, who acts on it, and why a person stays in charge. (AI for Logistics, 2026-10-31) - [Where an AI integration reads from: your system's API or a copy of its database](https://redfly.ai/insights/where-an-ai-integration-reads-from-your-systems-api-or-a-copy-of-its-database): An AI integration can read from your warehouse system's API or from a copy of its database. Here is what each path needs from your IT team and how we choose between them. (AI for Logistics, 2026-10-31) - [Where does AI fit in a warehouse? A way to find out without guessing](https://redfly.ai/insights/where-does-ai-fit-in-a-warehouse): You do not need to know what you want from AI. Start from how the work is done today, walk the floor, and look for the places where people re-key, re-check or wait. (AI for Logistics, 2026-10-15) - [Why AI pilots fail on a busy floor, and what a good one looks like](https://redfly.ai/insights/why-ai-pilots-fail-on-a-busy-floor): An AI demo runs on clean data in a quiet room. A floor has peaks, exceptions and shift changes. Here is why pilots fall over there, and how to run one that holds. (AI for Logistics, 2026-10-15) - [Is your data ready for AI? What a logistics operation actually needs](https://redfly.ai/insights/is-your-data-ready-for-ai): Your data is in several systems and several sites, and some of it is paperwork, video and sensor readings. Here is what ready means in practice, and what to do when it is not. (AI for Logistics, 2026-10-15) - [Cloud or onsite: where should AI run in a logistics operation?](https://redfly.ai/insights/cloud-or-onsite-where-should-ai-run): Some AI work can run in the cloud and some has to run at the site. The internet link, the floor and the decision itself tell you which, one integration at a time. (AI for Logistics, 2026-10-15) - [One integration at a time: how to start with AI without a big-bang project](https://redfly.ai/insights/one-integration-at-a-time): The way to bring AI into a logistics operation is one well-built integration, in seven steps worked through together, rather than a wide programme that never quite lands. (AI for Logistics, 2026-10-15) - [What to bring to a first conversation about AI](https://redfly.ai/insights/what-to-bring-to-a-first-conversation-about-ai): You do not need a wish list or technical knowledge. Bring how the work is done today, who does it and where people wait or re-key; that is enough for a useful first hour. (AI for Logistics, 2026-10-15) - [Can you trust what the AI tells you?](https://redfly.ai/insights/can-you-trust-what-the-ai-tells-you): AI answers can differ from one run to the next, which matters on a floor. Trust comes from a set of checks, a person in charge, and a pilot that earns it. (AI for Logistics, 2026-10-15) - [Will your people use it? Designing AI output for the people who act on it](https://redfly.ai/insights/will-your-people-use-it): Adoption is decided on the floor, not in the meeting room. AI output has to land where people already look, stay short, and be right often enough to be worth reading. (AI for Logistics, 2026-10-15) - [Spotty internet: what has to run at the site when the connection drops](https://redfly.ai/insights/spotty-internet-what-runs-at-the-site): Freezer dead zones and remote links mean the connection will drop. Some AI work can wait; some cannot. Decide per integration what runs at the site, and how it catches up. (AI for Logistics, 2026-10-15) - [A melting pot of systems: integrating AI when every site runs something different](https://redfly.ai/insights/a-melting-pot-of-systems): After a few acquisitions, every site runs its own warehouse system. One integration can read from each, through its API or a copy of its database, woven into the systems each site already runs. (AI for Logistics, 2026-10-15) - [Paperwork, video, audio and sensors: AI beyond the transaction table](https://redfly.ai/insights/paperwork-video-audio-and-sensors): Orders and stock are only part of what an operation knows. Emails, bills of lading, dock cameras, radio calls and sensors each suit different AI work, and each has its own meaning of ready. (AI for Logistics, 2026-10-15) - [Automated dock appointment booking, step by step](https://redfly.ai/insights/automated-dock-appointment-booking-step-by-step): How dock booking works today, what an AI integration reads, how it proposes slots, who confirms them in the redfly Portal, and how it is piloted before the dock team relies on it. (AI for Logistics, 2026-10-15) - [Adding a fourth warehouse: what it takes to bring a new site onto the central copy](https://redfly.ai/insights/adding-a-fourth-warehouse): Bringing a new warehouse onto a central copy of your data is real work done by people. Here is the honest setup list, and why each site after the first is easier. (Supply chain, 2026-10-10) - [From nightly exports to a central copy: what changes for a vendor's dashboard](https://redfly.ai/insights/from-nightly-exports-to-a-central-copy): When your dashboard reads a central copy kept current from every customer site, instead of files each site sends overnight, freshness, completeness and recovery all change in ways your users notice. (Remote sites, 2026-10-10) - [One product, many sites per customer](https://redfly.ai/insights/one-product-many-sites-per-customer): Vendors rarely sell one site to one customer. A separate space per site in one central store keeps each site apart and lets your application add a customer's sites together. (Remote sites, 2026-10-10) - [A support desk that can see the site](https://redfly.ai/insights/a-support-desk-that-can-see-the-site): When every deployment's current data reaches a central copy within a couple of minutes, your support team can look at a site before calling it and start from facts. (Remote sites, 2026-10-03) - [What your customer's IT team will ask before your service goes on their network](https://redfly.ai/insights/what-your-customers-it-team-will-ask-before-your-service-goes-on-their-network): Before a sync service runs inside a warehouse or plant, the IT team will ask what connects where, what is encrypted, and what runs on which machine. Here are honest answers. (Remote sites, 2026-10-03) - [Why a vendor should not build site-to-cloud sync twice](https://redfly.ai/insights/why-a-vendor-should-not-build-site-to-cloud-sync-twice): Vendors with technology at customer sites often rebuild data collection for each deployment. Here is what the second and third builds really cost, and what a supported connection changes. (Remote sites, 2026-10-03) - [Per-site encryption keys: why every warehouse gets its own](https://redfly.ai/insights/per-site-encryption-keys-why-every-warehouse-gets-its-own): Giving each warehouse its own encryption boundary limits what one leaked key can expose, keeps rotation a local job, and stops a single weak site from putting every other site at risk. (Sync, 2026-09-26) - [What "one IT contact" means in practice for a multi-site rollout](https://redfly.ai/insights/what-one-it-contact-means-in-practice-for-a-multi-site-rollout): A plain account of the single IT contact a multi-site data rollout needs: what that person does, where their time goes, and why one named owner moves faster than a committee. (Supply chain, 2026-09-26) - [Why nightly exports are the wrong shape for multi-site reporting](https://redfly.ai/insights/why-nightly-exports-are-the-wrong-shape-for-multi-site-reporting): A nightly export shows head office every site as it stood yesterday. We look at what that delay costs, and why a central copy current to within minutes changes the work. (Supply chain, 2026-09-26) - [How to centralize warehouse data from every site without replacing what runs the warehouse](https://redfly.ai/insights/centralize-warehouse-data-from-every-site): A small service beside each warehouse database sends changes outward to one central copy. The head office gets one view of every site; the warehouse floor keeps the system it already trusts. (Supply chain, 2026-09-25) - [Why reports should never run on the warehouse database](https://redfly.ai/insights/why-reports-should-never-run-on-the-warehouse-database): The operational database exists to run the floor. Every report that queries it competes with picking, packing and shipping. A central copy takes that load away and leaves headroom for peak. (Supply chain, 2026-09-25) - [What offline support really means when a warehouse loses its link](https://redfly.ai/insights/what-offline-support-really-means-for-a-warehouse): Changes queue on site, compressed and encrypted, and resume table by table when the link returns. Delivery is at-least-once, so nothing is lost. Here is exactly what is and is not promised. (Sync, 2026-09-25) - [One report across time zones: getting the clocks to agree](https://redfly.ai/insights/one-report-across-time-zones): Sites in Eastern, Central and Pacific time each keep their own clock. A central copy, stamped by site and current within minutes, gives head office one consistent day, shift or peak. (Supply chain, 2026-09-25) - [The true cost of a cloud copy of your warehouse data, and the cheaper shape](https://redfly.ai/insights/the-true-cost-of-a-cloud-copy-of-warehouse-data): A cloud relational database sized for every site is the expensive shape. Low-cost document storage, fed only with changes over ordinary internet links, is the cheaper one. Here is why the shapes differ. (Cloud economics, 2026-09-25) - [What a design partnership looks like for a logistics company: one site first](https://redfly.ai/insights/design-partnership-for-a-logistics-company-one-site-first): Who qualifies, what the first site involves, who drives the rollout, how sites and tables are added, and how the contract and subscription work. A plain account of partnering with redfly. (Design partners, 2026-09-25) - [Outbound only, encrypted per site: how warehouse data leaves the building safely](https://redfly.ai/insights/outbound-only-encrypted-per-site): The sync service opens outbound connections only, compresses and encrypts each site's changes with that site's own key, and keeps each site in its own central space. What a security reviewer will ask. (Security, 2026-09-25) - [Peak season for warehouses: why headroom matters more than horsepower](https://redfly.ai/insights/peak-season-headroom-matters-more-than-horsepower): On Black Friday the warehouse server has no room to spare. Moving reports off it, with a sync service that throttles itself when the site is busy, buys the headroom that keeps the floor steady. (Supply chain, 2026-09-25) - [Ten billion rows over the public internet: what we learned](https://redfly.ai/insights/ten-billion-rows-over-the-public-internet-what-we-learned): We have kept more than ten billion rows in step across ordinary public internet links, not dedicated lines. Here is what held up, what did not, and what surprised us. (Sync, 2026-09-25) - [IoT and fleet telemetry: high writes, flaky links, millisecond reads](https://redfly.ai/insights/iot-and-fleet-telemetry-high-writes-flaky-links-millisecond-reads): Devices and vehicles write constantly over links that come and go, while the operations screen needs the latest state in milliseconds. How we split the two jobs. (Sync, 2026-09-25) - [Drift, backfill and recovery: the unglamorous parts of sync](https://redfly.ai/insights/drift-backfill-and-recovery-the-unglamorous-parts-of-sync): Copies of data quietly disagree, restarts lose their place, and schemas change under you. What drift is, how it happens, and how an honest sync service detects and repairs it. (Sync, 2026-09-25) - [SQL Server change tracking: why we read changes on a short interval](https://redfly.ai/insights/sql-server-change-tracking-why-we-read-on-an-interval): SQL Server offers two ways to learn which rows changed. redfly uses the lighter one, change tracking, and reads it on a short interval. Here is why, and what it means for freshness. (Change capture, 2026-09-25) - [PostgreSQL logical replication as a cache feed](https://redfly.ai/insights/postgresql-logical-replication-as-a-cache-feed): PostgreSQL can publish every row change as a stream. redfly subscribes to that stream to keep Redis in step. What publications and replication slots are, and what to watch in operation. (Change capture, 2026-09-25) - [MongoDB change streams as a sync source](https://redfly.ai/insights/mongodb-change-streams-as-a-sync-source): MongoDB change streams deliver inserts, updates and deletes as they happen. redfly uses them to keep Redis, or a central store for site sync, in step, and to resume cleanly after an interruption. (Change capture, 2026-09-25) - [Transparent caching: why your codebase should not know a cache exists](https://redfly.ai/insights/transparent-caching-why-your-codebase-should-not-know-a-cache-exists): A cache your application code can see is a cache your team must maintain. We explain why reads should come from memory without the codebase ever knowing, and what that changes. (Caching, 2026-09-25) - [Joins on Redis: serving multi-field queries from memory](https://redfly.ai/insights/joins-on-redis-serving-multi-field-queries-from-memory): Most teams cache single keys in Redis and send every filtered or joined query back to the database. We explain how those queries can be served from memory too, and what that changes. (Redis, 2026-09-25) - [Cache invalidation with no invalidation code](https://redfly.ai/insights/cache-invalidation-with-no-invalidation-code): Stale-data bugs live in the invalidation calls scattered through a codebase. We explain how a sync service that watches the database removes the need to write them at all. (Caching, 2026-09-25) - [Automatic database fallback: reads that survive a cache outage](https://redfly.ai/insights/automatic-database-fallback-reads-that-survive-a-cache-outage): A cache that is down should make an application slower, not broken. We explain how reads fall back to the database automatically and why that matters more than hit rate. (Redis, 2026-09-25) - [Surviving Black Friday with the database at 99% CPU](https://redfly.ai/insights/surviving-black-friday-with-the-database-at-99-percent-cpu): What a retailer's database does on its worst day, why product and account pages fail before orders do, and how reads served from memory keep the site up without a bigger database. (Scaling