Global market data infrastructure

One feed in.
Every market covered.

Send us your own price feed. AlphaFlowing Pricing Hub turns it into a complete candle history and delivers it — live and historical — from a node close to every client you have. Your prices stay yours, and so does your data: we never re-price it, and we never resell it.

Want to see it first? We run a demo environment with sample market data you can click through.

PRICING CENTER
Pricing center Regional node Illustrative — nodes are placed to match your client base
Your feed stays yours
Your price source comes in. We change nothing, and resell to no one.
A node where your clients are
Placed to match your markets. Entering a new region never means rebuilding.
Nobody has to watch it
Storage, segmentation, cleanup and scaling are handled by the service itself.
Candle queries in 2 ms
Measured median at 400 QPS. P95 7.5 ms, 99.997% of requests served. Your clients' round-trip time to the node adds to this.

The problem

You already have the prices. Your clients don't have the experience.

The feed was never the hard part. Turning it into charts that open instantly and scroll back forever — in every market, at the same speed — is. If the prices are yours and your clients are not all in one place, this is your problem too, whatever you call your firm.

01

A feed is not a chart

What you have is a stream of ticks. What clients want is a full history at every timeframe. Between the two sits a storage and aggregation project that has nothing to do with your trading business — and still costs you a team.

02

Storage only grows

Ticks arrive every day and never leave. Segmentation, compression, cleanup, expansion — the day nobody is watching is the day it breaks, and it usually breaks at the open.

03

Overseas clients get the slow version

Your servers sit at headquarters. Every chart your new-market clients open takes longer than it should. The traffic you paid for stalls on market data.

What it does

Your feed goes in. Your clients, everywhere, get it out.

AlphaFlowing Pricing Hub does not supply market data. You connect your own price source; Pricing Hub stores it as a complete candle history and delivers it from a node near your clients.

01 · Historical candles

However far back they scroll, it's there

Once your feed is connected, every timeframe is aggregated and stored for you.

  • Minute to monthly, aggregated automatically — you never write that logic again
  • Existing history can be bulk-imported, so the chart doesn't start on the day you signed
  • Enough bars in one request that the chart opens full — and keeps going when they scroll
  • The in-progress bar updates live; nobody waits for the close
What clients notice: charts open instantly, scroll smoothly, and switching timeframes reloads nothing.
1.09305 1.0980 1.0890 1.0860 09:00 10:00 11:00 12:00 EURUSD · M1
02 · Live distribution

One feed, every market

Your quotes are pushed to clients worldwide exactly as you sent them. AlphaFlowing Pricing Hub does not price, and does not touch the numbers.

  • Continuous push — clients never poll
  • Same source as the candle history, so the last bar always meets the live price
  • Automatic reconnect, with connection state always explicit to the client
  • Market closed, no quote, connection lost — each a distinct state, never a frozen number pretending to tick
What clients notice: quotes that hold steady, don't jump, and don't quietly die — the price they see is the price you sent.
SYMBOL BID ASK EURUSD 1.09418 1.09424 XAUUSD 2318.44 2318.71 GBPUSD 1.27106 1.27114 USDJPY 157.284 157.291 BTCUSD 67 412.5 67 419.0 USOUSD 78.412 78.437

Lower cost

Building this yourself costs more than the budget line says

Storage, aggregation, distribution, scaling — none of them is hard on its own. Together they are a team for a year, and then somebody has to keep feeding them.

Build it yourself

Ongoing spend, rising every year

Storage Segmentation, compression, cleanup, scaling — all written and maintained by you
Aggregation Candle generation and consistency across every timeframe, implemented in-house
Global reach Rebuild the stack per market, then keep them in sync yourself
Headcount At least one dedicated engineer, plus an on-call rotation
Time to live Measured in quarters
Feed ownership Yours
Cost curve Rises with data volume and market count
Use Pricing Hub

Connect once, pay for what you use

Storage Already there, including segmentation, compression and disk protection
Aggregation Generated automatically as the feed arrives
Global reach Switch on the regional node you need
Headcount No dedicated engineer, no on-call
Time to live Measured in days
Feed ownership Still yours — we don't price it and don't resell it
Cost curve Tracks actual usage, and stays predictable

There is of course a third option: buy a feed from a data vendor. That is easier — but what you give up is your own price. Your markup, your client tiers, your quoting strategy, all gone. AlphaFlowing Pricing Hub is the one path that spares you the build without costing you your prices.

Performance & assurance

Market data proves its worth on the bad days

Two kinds of number below, and they are not the same kind. First what a node was observed doing on a public test. Then what we owe you contractually, backed by service credits if we miss it.

Observed What a node did serving candle queries Public regional node · 1 shared vCPU · EPYC 7313
Median response
2ms
P95 response
7.5ms
Request success rate
99.997%
Sustained load
400QPS

These are the node's own response times — a client also waits for their round-trip to the node. k6 over the public internet, London to London, from another instance at the same provider. The vCPU is one logical thread, not a dedicated core. Ten instruments, two timeframes, 200 bars per request with the in-progress bar, gzip on, 400 QPS for 600 s: 7 admission rejections in 240,001 requests.

This is why the node belongs next to the client

What a client waits for is our part plus the distance. Our part is fixed and small. The distance is whatever sits between your client and the nearest node — and on a different continent it is the whole of the wait.

2ms
Candle query, measured
+RTT
Client to node, geography
Observed What a node did pushing live quotes Public regional node · 1 shared vCPU · EPYC 7313
Quotes pushed in two minutes
1.25million
One price change reaches
700clients
Clients on one node
2,000live at once
Arrived late or out of order
0not one

A client is never shown a price you have already replaced — a superseded quote replaces the old one rather than queueing behind it.

Committed What we owe you, in writing Service agreement · backed by credits
Monthly service availability
99.9%
Recovery time
4hours
Support response
4business hours

These are the Production commitments, included at no extra cost. Pilot runs on best effort with no availability commitment. Premium SLA lifts them to 99.99%, one-hour recovery and 30-minute response for US$350 a month, and is included with Scale — see plans.

* How availability is measured, what is excluded, and the service credits payable when a target is missed.

A full disk won't take the service down

As storage approaches its limit the service reclaims space on its own and protects the newest data first, instead of dying when the volume fills. It is the most common way a market data service fails; we built the answer into the service.

One market's peak doesn't slow another

Resources are allocated across regional nodes by actual load. The London open does not make your Sydney clients wait.

Nodes stay in step on their own

Differences between a regional node and the center are detected and filled automatically. No manual syncing, and no region quietly missing a stretch of history.

Easy integration

No re-architecture, no new team

Your system almost certainly emits FIX quotes already. The rest is connecting it, and pointing your clients at two endpoints.

STEP 01

Connect your feed

Point your existing quote output at our receiver. Standard FIX — most trading systems support it out of the box, and nothing needs rebuilding for us.

Day one
STEP 02

Import history (optional)

Existing candle history can be bulk-loaded so there is no gap. No history is fine too — it accumulates from the day you connect.

Depends on volume
STEP 03

Point clients at two endpoints

One pulls candles, one subscribes to quotes. We supply working examples and sit with your engineers through integration.

A few days

Nothing else changes

Only the storage and distribution of market data is replaced. Pricing, accounts, orders, risk and back office stay exactly as they are — we cannot reach them.

No lock-in

The data is yours and the interfaces are ordinary requests and pushes. Switch back whenever you like — we would rather keep you on service quality than on migration cost.

A person, not a ticket queue

Someone works directly with your engineers during integration, and the same person stays with you after go-live.

Delivery

Three ways to connect, depending on your compliance

From fastest to live, all the way to nothing ever leaving your own network. Hosted nodes come with every plan; self-hosted and hybrid are part of Scale, because running inside your network is a different piece of work.

Fastest

Hosted nodes

Your quotes go to our regional nodes; clients read from the closest one.

Inside your network Your price feed
Run by us Pricing centerRegional nodesStorage

Your feed leaves your network; everything downstream is ours to run.

  • Nothing to deploy
  • No operations, no on-call
  • Right for evaluation and for most clients
Stays in your network

Self-hosted

The whole stack runs in your own data center or cloud account.

Inside your network Your price feedPricing centerRegional nodesStorage
Run by us

Nothing crosses the boundary. We deliver, upgrade and support it remotely.

  • Neither quotes nor history leave your network
  • Meets strict data-residency requirements
  • We handle delivery, upgrades and remote support
Grows by market

Hybrid

You keep the core; regional nodes follow your client base.

Inside your network Your price feedPricing centerStorage
Run by us Regional nodes

History and the center stay with you; reach comes from our nodes.

  • Data sovereignty stays with you, reach comes from us
  • Add nodes as you enter new markets
  • Or the other way round, if that suits you better

Pricing

Your client count is not our business model

Market data is nearly always sold by the seat or by the call, so growing your book means growing your bill. Not here. Plans differ by how many regions you cover and what you need us to commit to — never by how well you are doing.

Pilot 3 months free

For running your own feed through it for real, before you commit to anything.

Freefor three months

Then US$500 a month, or A$700 invoiced in AUD

  • 1 regional node
  • Up to 400 sustained requests per second
  • Up to 4,000 concurrent streaming clients
  • Full candle history and live quotes, same as every plan
  • Hosted, not self-hosted
  • No availability commitment — best effort

Additional nodes US$100 a month. When the three months are up, stay on Pilot as long as it suits you — there is no deadline to move.

Production Most brokers

For a book you are actually running, in the markets you actually sell into.

US$1,000/ month

A$1,400 if you prefer to be invoiced in AUD

  • 2 regional nodes
  • Up to 400 sustained requests per second on each
  • Up to 4,000 concurrent streaming clients on each
  • 99.9% monthly availability, 4-hour recovery
  • 4 business hours to a service-affecting report
  • Hosted nodes

Additional nodes US$100 a month. Premium SLA — 99.99%, 1-hour recovery, 30-minute response — US$350 a month.

Scale Advanced

For data residency rules, dedicated capacity, or anything that does not fit a list price.

Custom

Priced against your actual shape

  • Self-hosted and hybrid delivery
  • Dedicated cores, capacity sized to your load
  • Premium SLA included
  • Nodes as your markets require
  • A named contact, not a queue

Self-hosting inside your own network starts here — it is the only plan that leaves our infrastructure.

True on every plan

No per-user fee Five hundred clients or five thousand, the invoice is the same.
No per-request fee Charts exist to be opened. Nobody here is counting them.
No data margin The feed is yours. We do not resell it, least of all back to you.

What you are really comparing this to. Not another vendor — the engineer you would hire to build this and then keep it running, plus the on-call roster behind them. Production is US$12,000 a year.

Prices are inclusive; nothing is added at invoicing. USD and AUD are separate list prices rather than a conversion of one another, so neither moves when the exchange rate does. Billing is monthly, in advance.

See the demo first. Then send us your own feed.

We run a demo environment with sample market data — charts, live quotes, timeframe switching, all clickable. If it looks right, connect your own price source for a trial, and what's on the chart is already your price.

Three months free on Pilot, then US$500 a month. No per-user and no per-request charge on any plan, and no quote to wait for.