DeFi Orbit vs. building it in-house
Building your own position data pipeline looks like a few weeks of work. It is a system you will maintain for as long as you need the data. We know, because we run one every day. This page sets out what it takes.
What each path costs you
What "we will just index it ourselves" turns into
Reading a balance is one call. A position you can report on needs every piece on this list, for every protocol, kept running. This is the work we do full time.
- 01An indexer per protocolContracts, events and edge cases differ for each one. None of the work carries over.
- 02Event decoding and storageDeposits, withdrawals, fee collections and rebalances, in order, with reorgs handled.
- 03Historical price dataA price for every token at the time of every event. Today's price is the easy part.
- 04A calculation enginePnL against holding, fees, impermanent loss, APR and health factors, each with rules you have to decide and document.
- 05Snapshot jobsDaily values for every position, forever, so you can answer what it was worth on any date.
- 06ReconciliationChecks that your numbers still match chain state. Without them you find out from a client.
- 07Monitoring and on-callRPC failures, reorgs, stuck jobs and rate limits, at any hour.
- 08Then the protocol upgradesA new version or a migrated contract, and part of the list starts over.
The costs that are not in the build estimate
The first version is the cheap part. This is what the estimate leaves out.
It is never finished
A build estimate covers getting to the first number. It does not cover the years of upgrades, new deployments and new chains after it.
It depends on one or two people
The engineers who understand the pipeline become the only ones who can fix it. When one leaves, so does that knowledge.
Wrong numbers cost more than no numbers
A PnL figure that is slightly off goes into an investor report before anyone notices.
The bill is not predictable
RPC usage jumps with a backfill, a busy chain or a new client. So does the invoice.
You have to defend every number
Auditors and investors ask how a figure was calculated. With your own pipeline, your team writes and maintains that answer.
It competes with your roadmap
Every week on indexers is a week not spent on the product your customers pay for.
The short list of reasons to build
There are three, and they apply to few teams. If one is yours, we will tell you so.
The data is your product
If position analytics is what you sell, owning the pipeline can be worth it. Expect a standing team, not a project.
You need something we do not cover
Check supported networks and protocols first. If it is missing, ask about Custom Data: adding it is usually faster than you building it.
You have strict data residency rules
DeFi Orbit is a hosted API, so if your policy rules out any external service, build in-house. Before you decide, note that we only process public on-chain data for the addresses you query, with no keys, funds or personal data, which satisfies most residency reviews.
Questions teams ask before deciding
Yes, and that’s a sensible order. There’s no subscription or commitment on credit packs, so you can ship with our data now and build your own pipeline later if it still makes sense. Having correct numbers to compare against makes an in-house build far less risky than starting from nothing.
No. Run the API next to your own pipeline on the same wallets for a couple of weeks. Where the numbers agree, you’ve validated your work; where they differ, you’ve found an edge case before a client did. Then you can decide whether to keep going or hand the rest to us.
Test them on wallets you already know. Every figure is calculated from on-chain events using documented methods for PnL, impermanent loss, fees and APR, so you can trace any number back to the transactions behind it. If something doesn’t match your expectation, tell us and we’ll show you the calculation.
Each call costs a fixed 1 to 15 credits, with packs starting at 500 credits for $10, so you can price your usage before you write a line of code. Building means engineer time, RPC bills that grow with every wallet and chain, and someone on call when a protocol upgrades. For most teams, a month of our credits costs less than a day of that.
We add protocols and networks as scoped Custom Data projects, with a timeline and a price agreed before we start. Once live, the new protocol works everywhere: the API, DeFi Orbit AI and Portfolio. It’s usually faster than building the indexer yourselves, because the calculation layer already exists.
No. Everything is read-only and works from public wallet addresses. There’s nothing to connect and nothing we could move, which also means the integration adds no security surface to your own systems.
Find out in an afternoon, not a quarter
Run one API key against wallets you know before you commit a team to building it.