Monday: Multi-inputs, mining, and minimal mode

This week's release includes more features to help you answer tickets faster.

See more about your searches

We now highlight different types of transactions:

  • Multi-input transactions (e.g. when a customer sends two deposits at once, and the system only 'sees' one)

  • Mining transactions (e.g. when a customer mines to their exchange address)

  • Token & asset transfers (including what blockchain they "live" on, e.g: (this allows you to quickly check the coin type in cases of swaps/hard forks)

  • Smart contract or complex blockchains (like ETH, EOS, XRP, XLM) that have failed during sending, e.g.

Customisations:

Introducing minimal mode! Search your transactions without distractions.

Before:

After:

About txBatch v2

Hi, I'm James, developer at txBatch. From mid 2017 to 2018, I worked at the #20 cryptocurrency exchange.

Like many exchanges, we were based on dreams - "growing into a truly kickass platform" - and we grew at a staggering pace in the bull run, from 10,000 to 2 million users. All of a sudden, it was 2018 and we had 500+ tickets per day with a 10 person support team.

Of course, we couldn't keep up with demand. Our online reputation tanked - customers were waiting 30+ days for answers. Me and the other support employees were coming into work bored and frustrated because no matter how hard we tried, they just kept on stacking up.

There was one bottleneck in particular that was causing 80% of the holdup/mistakes. It was manually checking block explorers (websites used to query blockchain data).

You see, we had to find very specific data when replying to our customers. About half our questions were asking about a missing deposit or withdrawal, and they generally took 7 mins to resolve. Which stacked up.

Every day, from clock in to clock out, 50% of our day would be spent searching transactions. Manually, individually, over and over.

And we didn't make a dent in our tickets for months, even after doubling our headcount.

We made typos and mistakes in support answers, due to improperly copied details. For those of us with higher responsibilities, there was additional risk. One time I credited a user $15,000 more than they should have had by mis-copying the decimal values from the popular Ethereum block explorer Etherscan. Luckily I reversed it, but I never forgot that brief moment of terror.

(Example of how $15 can turn into $15,000) Windows usually detects the full-stop/period (.) as an indication of a sentence stopping. Because of this, it becomes very easy to only copy half of a number, and tell a customer (or, credit them) a different amount to what they deposited. When exchanges are busy and handling hundreds of tickets per day, the risk of errors like these increase exponentially.

I initially tried to build a solution like txBatch at our exchange, but quickly realised that it could be useful to a lot more people than us. So in October 2018, I co-founded txBatch with my friend Winston.

We got to work identifying exactly what the common issues were between exchanges, and produced an initial 'bulk search block explorer' prototype in November. Our v1 was out in December, and we began sending it around the industry.

Here was our demo, at the time: txBatch 5 minute demotxBatch 5 minute demo

We got a lot of positive interest, which was great, because it meant we'd found a real problem. Unfortunately, the v1 was full of bugs and was relying on a fickle data stream of public block explorer APIs. We needed a more professional solution.

So, for the last 3 months, we've been working on txBatch v2, and have finished rewriting our back-end.

Here's a run through of the biggest improvements.

#1: Significantly more coins, tokens, and digital assets

By far our biggest request has been to increase the amount of blockchains we support. We started off with BTC, ETH, ERC20 tokens and USDT, but since the crypto markets are rather diverse these days (and since a lot of deposit/withdrawal issues occur with lesser-known altcoins), we needed to support more.

We now support hundreds of popular cryptocurrencies, and have support for tokens on networks such as ETH, XLM, XRP, XEM, and NEO.

The data we get from each blockchain is also much more specialised to that blockchain, i.e. Ethereum smart-contract transactions that have run out of gas will display as such.

Snapshot of our coin support, based the OnChainFX (Messari) rankings report. Going off the current CoinMarketCap numbers, we now support more than 80% of the top 100 cryptocurrencies by market cap.

#2: Customisable display

You can now customise what headers you see on your transaction bulk search table, how transaction data is displayed, and the block explorer links for transactions.

You can also export results to Excel with a handy "Export" button, allowing you to do more granular data manipulation.

Note: The data you can export in fact contains more than you typically see in the interface. For example block heights, and smart contract addresses.

#3: Faster batch searches

The biggest improvement is our completely rewritten back-end. It is significantly more secure, private, and fast.

50+ transactions can now be searched across blockchains, in less than 3 seconds. Take that, monster ticket with 250 transaction IDs!

We achieved this by changing our back-end to accept batch requests. Our previous version searched up single transactions in order, spaced out by about 1 second to prevent memory issues.

Our new back-end is hooked up to a diverse set of high-powered blockchain nodes and can search concurrently (at the same time) for near-instant results.

We operate a variety of blockchain nodes and have partnered with many professional services including AnyBlock's Eth.Events, EOSFlare.io, and MyCrypto's MoneroVision.

What this means for our users is that they can search up transactions faster and more securely than using traditional block explorers, and reply to traders faster.

#4: Increased security + privacy

There are those that may argue that publicly available blockchain data doesn't carry any kind of security risk.

Well, we disagree. When paired with the knowledge of who is searching that blockchain data, then that data becomes very risky in the wrong hands.

So we take it out of ours. For the privacy conscious, txBatch...

  • is now HTTPS-only
  • handles searches in an isolated environment to user data
  • handles searches anonymously (all that our nodes "see" is that txBatch is asking for transaction ID XXXYYYZZZ)
  • never uploads anything to our servers without your consent (not even your display preferences. They're stored locally until you choose to add them to your account)
  • allows users to view, update, and manage their data at any time
  • only logs how many searches you do per day, reset every day
  • stores transaction results on your local browser, never on our servers

New txBatch features such as address tagging ("name" addresses as coming from a particular exchange or wallet) will be end-to-end encrypted, and require authentication to store.

You'll be able to use zero-knowledge login provider Blockstack.org or opt for a traditional password. (And that password, too, will be encrypted before being uploaded to our servers, never stored as plain text.)


We'll be posting more over the coming weeks and select testers will be invited to join our premium beta.

Eventually, we'll be launching a freemium version of txBatch, where smaller players (such as businesses that accept crypto payments) can get 30 free transaction searches per day. Data privacy is very important to us, so there will be a variety of sign-in methods including Telegram, WeChat, Blockstack, and Signal.

Our focus going forward is to create the best customer support tool in crypto, and we appreciate everybody's help so far.

No published changelogs yet.

Surely txBatch will start publishing changelogs very soon.

Check out our other public changelogs: Buffer, Mention, Respond by Buffer, JSFiddle, Olark, Droplr, Piwik Pro, Prott, Ustream, ViralSweep, StartupThreads, Userlike, Unixstickers, Survicate, Envoy, Gmelius, CodeTree