Skip to content
SIH Buddyby Ganeev Singh
Dev

🔥 Roast My Pick · SIH26146

AI-Powered Monitoring & Analysis of Bitcoin Transaction Traffic

National Technical Research Organisation (NTRO)

Mild17/100

Good pick. Genuinely. Now sit down, because the judges are going to try anyway — and this is what they will try.

Strong pick. The dataset is provided and the specification is complete, so this is genuinely buildable — make the network-plus-blockchain correlation your distinctive angle, and treat the explainability as core since a lead an investigator cannot understand is useless. Roughly 240–500 teams are expected to go here.

The receipts

Every red flag on this statement, in full. These are the four places it bites.

  1. Exhibit A

    The network-blockchain correlation only exists because the synthetic dataset pairs IPs with transactions — in reality that linkage is rarely available, so acknowledge that your capability depends on the provided data model

  2. It gets worse

    Address clustering heuristics produce false merges, and over-clustering wrongly implicates unrelated parties, so cluster precision matters in a forensic context

  3. Still reading?

    The description requires a trained model rather than rules, so a purely heuristic submission does not meet the stated requirement

  4. And the finisher

    Flagging a wallet as suspicious is an investigative accusation, so explainability and confidence must be genuine rather than decorative

The damage report

Every score this statement earned, and what each one actually costs you.

  • Feasibility

    4/5

    Actually buildable, which on this slate is rarer than it sounds. Do not squander it on scope.

    The description states a synthetic dataset will be provided modelled on real Bitcoin fields, address clustering heuristics are well-documented, graph analysis and anomaly detection are standard, and everything runs offline on Linux with no live blockchain access needed.

  • Innovation scope

    4/5

    There is something genuinely new here. Do not bury it under another dashboard.

    How you fuse the network-layer IP and timing data with the blockchain-layer wallet graph is the genuinely open part — most public blockchain analysis uses only on-chain data, so correlating it with network metadata is a real and distinctive contribution.

  • Clarity

    5/5

    The ask is unambiguous, which quietly removes your favourite excuse.

    The description lists the exact input fields, the required graph construction, mandates a trained model rather than rules, specifies ranked explainable alerts with confidence, and even defines the synthetic dataset schema — leaving nothing ambiguous.

  • Acceptance potential

    4/5

    Strong footing before you have written a line. Try not to waste it.

    A strong pick — the dataset is provided, the specification is complete, and the network-plus-blockchain correlation is a genuinely distinctive angle that separates this from the many purely on-chain analysis projects, with NTRO's investigative framing giving it clear purpose.

  • Effort

    Heavy

    Heavy. Somebody on this team is not sleeping in week three. Pick who, on purpose.

    Ingestion, entity graph construction, the ML detection model, explainability and a link-analysis visualisation are five connected pieces, though each rests on established techniques.

  • Demo-ability

    Easy

    Easy to demo — and so is everyone else's. Working is the floor here, not the achievement.

    A link-analysis graph showing wallets clustering into an entity with a fund-flow trail highlighted is a compelling, investigative-looking demo that reads instantly.

  • Data

    None supplied

    No dataset comes with this one, so every accuracy figure you quote is a number about labels you invented.

    Nothing is provided with the statement. You are sourcing, cleaning and labelling it yourself, and that work is invisible in the demo but very visible in the questions.

The demo they will have already seen

Somewhere around 240–500 teams are heading here, and the description is doing the choosing for most of them. They will read the same brief, reach the same architecture, and build a version of the same demo you are planning. Being correct is the floor. If your five minutes could be swapped with the team before you and nobody in the room would notice, you have not picked badly — you have built predictably, which costs exactly the same and hurts more.

What survives

The ground worth standing on when the questions start.

  • The synthetic dataset is provided with a defined schema, so there is no data-sourcing risk and everyone works from the same fields
  • Correlating network-layer IP and timing with the on-chain wallet graph is a distinctive angle that most blockchain analysis omits entirely
  • The common-input-ownership clustering heuristic is well-documented, so your entity clustering rests on established forensic technique

Nothing here is fatal. It is just the list of places this statement pushes back, and you now get to push there first.

Full analysis ↗Roast another

The framing is a joke. The findings are not — they are the same analysis on the statement page, and every line above is attached to a score or a fact in the record. It is one opinion with its reasoning attached, so argue with it before you trust it.