Custom data tools

Every delivery is fully documented: what it does, what it does not do, and how to run it. Fixed price and scope agreed in writing before work starts. All communication is written and asynchronous.

How it works

1. 1

   Describe what you need

   One email with the source, the output you want, and any constraints.

2. 2

   Receive a fixed quote

   Price, scope, and delivery window in writing. No hourly billing.

3. 3

   Receive the build

   The working tool plus complete documentation.

What I build

• Custom data extract

  A slice of the Instrumetriq Tier 3 archive: your coins, your date range, your fields, as parquet or CSV with a field dictionary. Seven months of constant schema (December 2025 to August 2026) across 250+ Binance pairs.

  Example: all DOGE and SHIB sessions for March 2026, sentiment and price fields only, as a single CSV.

  Source documented in three published research notes with DOIs.

  Details

  What it is

  A filtered slice of the Instrumetriq Tier 3 archive, cut to your specification. The archive covers 250+ Binance pairs at ~2-hour session granularity: market microstructure (price, spread, order book depth, funding, open interest) and X (Twitter) attention measurements (post counts, distinct authors, engagement, tone scores) per coin per session, plus ~700 intra-session price samples at ~10-second spacing.

  What you provide

  The coins, the date range, the fields you want (or 'all fields'), the output format (parquet or CSV), and whether you want all sessions or only sessions with observable social activity.

  What you receive

  The data files, a field dictionary describing every included column, and a coverage note stating the row count, the applied filters, and any gaps inside your requested range. Delivered as a download link.

  from EUR 90 delivered within 48 hours

  Order

• Custom scraper or monitor

  A scheduled scraper for a publicly accessible source you choose: pages, feeds, or APIs. Delivered as a documented script or a deployable service, with output as JSON, CSV, or parquet on your schedule.

  The Instrumetriq pipeline collected from X continuously for seven months across 250+ tracked assets at ~30-minute cycles.

  Details

  What it is

  A scheduled collector for a publicly accessible source you choose: web pages, feeds, or APIs. Built to run on your machine or server on the schedule you set, with deduplication and failure handling.

  What you provide

  The source (URLs or API), the data points to capture, the collection schedule, the output format (JSON, CSV, or parquet), and where output should land (local files or a storage target you control).

  What you receive

  The documented source code, a configuration file, run and scheduling instructions for your environment, a sample of captured output, and a written specification of behavior including rate limits and failure handling.

  EUR 250 delivered within 7 days

  Order

• Telegram or Discord alert bot

  A bot that posts when your condition fires on your data source. You define the trigger in plain language. Delivery includes the bot, hosting instructions, and a written specification of its behavior.

  The Instrumetriq extension's alert layer monitored 250+ assets against per-coin baselines.

  Details

  What it is

  The bot checks your source on a fixed interval and posts on the transition into your condition, not on every check, so a condition that stays true does not repeat itself into the channel. It reports and nothing else: it places no orders and writes nothing back to your source.

  What you provide

  The data source (a public API, feed, or page), the trigger condition in plain language, the destination (Telegram chat or channel, or a Discord webhook), and the message content you want per alert.

  What you receive

  The bot source code and configuration, hosting and setup instructions, a sample alert, and a written specification stating exactly when the bot posts and when it stays silent.

  EUR 300 delivered within 7 days

  See an example Order

• Scrape-to-dashboard build

  A pipeline from a source you choose to a browsable dashboard: collection, storage, and a static or lightweight web view of the results, rebuilt on the schedule you set.

  The Instrumetriq pulse page is built this way from the same pipeline.

  Details

  What it is

  The dashboard reads from stored history rather than querying the source on each visit, so views can show change over time and stay readable while the source is briefly unavailable. The Instrumetriq pulse page is an example of this shape, built from the same pipeline that produced the archive.

  What you provide

  The source, the metrics and views you want on the dashboard, the refresh cadence, and whether the dashboard should be a static page rebuilt on a schedule or a lightweight server.

  What you receive

  The collector, the storage layer, the dashboard, documentation for each part, and deployment instructions for your hosting.

  EUR 550 delivered within 10 days

  See an example Order

• Browser extension build

  A Chrome or Firefox extension built to spec: data display, page augmentation, or monitoring surfaces. Delivered as source plus a packaged build per browser, with store listing preparation if you want it.

  The Instrumetriq extension shipped on both stores with a free and a paid tier.

  Details

  What it is

  Built on Manifest V3 with the narrowest permission set the behavior allows, since permission scope is what store review questions most often. Publication happens on your own developer account, and review timing is decided by the store, not by the delivery window above. The Instrumetriq extension shipped on both stores with a free and a paid tier and is the reference build.

  What you provide

  The target sites or pages, the desired behavior in plain language, the browsers to support, and whether you want store publication prepared (listing text and assets) for your own developer account.

  What you receive

  The extension source code, a packaged build per target browser, store listing text and asset preparation if requested, and documentation of behavior and permissions.

  EUR 700 delivered within 14 days

  See an example Order

• Custom data analysis

  A written analysis of a dataset you provide or a public source you name, delivered with a reproducible notebook. Every number in the report is regenerated by the notebook from the input data.

  Reference builds: three published research notes with DOIs, each with a reproduction notebook.

  Details

  What it is

  A defined set of questions answered against your data: distributions, associations, comparisons, or the evaluation of a claim. The deliverable states what was measured, how, and what was and was not found. Negative results are reported as results.

  What you provide

  The data (file or public source), the questions you want answered, and any thresholds or definitions that matter to you.

  What you receive

  A written report, a runnable notebook that reproduces every figure and table from the input data, and a field-level description of any transformations applied.

  EUR 400 delivered within 10 days

  See an example Order

Order or ask

Order by form or by email. Either way you get a written reply.

Scope and terms

Scope

Fixed scope only. Changes after agreement, including changes of preference outside the written scope, are a new quote.

Sources

Sources must be publicly accessible. No authentication bypass.

Not offered

Authentication systems, payment processing, anything medical, anything requiring live calls.

Payment

Payment terms and invoicing are confirmed in writing before work starts.

Ownership

Delivered code and outputs belong to the buyer on full payment. I retain the right to reuse generic, non-client-specific components.

Acceptance

The delivery is measured against the written scope. If the delivery does not match the agreed scope, it is corrected at no cost.

Revisions

One revision round for scope-matching corrections is included. Support beyond delivery is not included unless quoted.