Telecommunications · fraud and network forensics

The same fraud comes back with new numbers.

Fraudsters change numbers, routes and accounts. The method stays the same. Synsa remembers every case your fraud and network teams have closed, and recognises the method when it returns.

What your teams live with

  • Fraudsters change numbers, routes and accounts. The method stays the same.
  • Investigators rebuild every case by hand from call records and tickets.
  • Outages repeat across regions, and each team finds the root cause again.

Two records, one shape

Illustrative example
Case 2024-118 · Fraud · closed
Night burst of calls to premium ranges from new prepaid SIMs. Blocked by range bar and SIM suspension.
NewCase 2026-041 · Alert
Spike in international calls from SIMs activated this week, 02:00.
Same shape New accountsNight burstPremium ranges

Synsa brings back case 2024-118 and what stopped it, before the losses grow.

Where Synsa reads from

  • Fraud management cases (Subex, Mobileum)
  • CDR and xDR summaries
  • SIM and subscriber lifecycle events
  • Trouble tickets (ServiceNow, BMC Helix)
  • NOC alarm history and incident reviews
  • SIEM logs

What you measure

  • Mean time to investigate a case
  • Fraud loss per incident
  • Share of cases matched to a known pattern
  • Repeat outages and mean time to restore
  • Analyst hours per case

Estimate your return

What repeat cases cost your analysts today.

420
analyst hours freed each month
3.0
full-time analysts' worth of work
€352,800
per year, before any fraud loss avoided

Estimate only, using your numbers. Assumes a case matched to an earlier one is resolved in half the time, and 140 working hours per analyst per month. In the first week we measure your real figures and agree the baseline with you.

From pilot to payback

Six weeks in your network. Then it keeps running.

  1. Week 1

    Connect

    Read-only links to fraud cases, tickets and CDR summaries, on your hardware. Baseline KPIs agreed and signed.

  2. Weeks 2–3

    Learn

    Past cases become Traces. Synsa finds the patterns that repeat. Your analysts confirm them.

  3. Weeks 4–6

    Run

    Every new alert is checked against confirmed patterns. Analysts see the earlier case and what stopped it.

  4. Week 6

    Prove

    Return-on-investment readout against the baseline, signed by both sides. The memory keeps running.

Context: telecom fraud cost operators about $38.95B in 2023, 2.5% of revenue (CFCA). At least 50% of generative AI projects were abandoned after proof of concept (Gartner, 2026).

Questions we hear

How does Synsa recognise telecom fraud that returns under new numbers?
Each closed case becomes a Trace built around its method: account age, timing, destinations and what stopped it. New alerts are matched on that method, so new numbers, routes or accounts still bring back the earlier case.
Which fraud and network systems does Synsa read?
Fraud management cases from Subex or Mobileum, CDR and xDR summaries, SIM and subscriber lifecycle events, trouble tickets from ServiceNow or BMC Helix, NOC alarm history and incident reviews, and SIEM logs. Access is read-only.
Do call records leave our network?
No. Synsa runs on your hardware, read-only, without internet. Subscriber data is masked on arrival.
How does Synsa fit with our fraud management system?
Synsa works alongside it. It reads closed cases from your fraud management system and shows analysts the earlier case behind a new alert, with what stopped it.
What is a Trace?
A Trace is one thing your company has learned: the shape of a problem, its facts, its fix and who confirmed it. Synsa uses Traces to recognise a problem when it comes back.

Bring one fraud type or one region. We'll show you what repeats.

Request a pilot

Cookie settings

Choose which cookies we may use. You can change this at any time through “Cookie settings” in the footer. Cookie policy