The SimoN p oject leve ages statistics combi ed with AI a d Machi e Lea i g (ML) to ge e ate highly ealistic, sy thetic households a d tax data. This capability allows the IRS to pe fo m igo ous e te p ise-wide systems testi g without copyi g o exposi g se sitive Fede al Tax I fo matio (FTI) f om p oductio systems.
I Phase 4, Skyma tics successfully e ha ced SimoN to ge e ate sophisticated data models that simulate o -complia t a d pote tially f audule t activities. I Phase 5, Skyma tics will expa d the simulatio e gi e eve fu the to model complex eal-life eve ts fo both i dividual taxpaye s a d busi ess e tities. The e ha ced SimoN e gi e will ge e ate comp ehe sive tax-specific data, i cludi g Mode ized e-File (MeF) XML tax etu s a d i fo matio etu s, which will feed di ectly i to dow st eam IRS pipeli es like the I dividual Maste File (IMF) a d Busi ess Maste File (BMF).
C ucially, the expa sio will adva ce the system’s ability to t ack a d simulate o -complia t taxpaye s, u de epo ti g, a d f audule t activities like ide tity theft, e abli g the IRS to p oactively test thei f aud detectio a d e te p ise systems agai st ealistic th eat sce a ios. The e gi e will co ti ue to be deployed a d ope atio alized withi the T easu y Wo kplace Commu ity Cloud (WC2) Platfo m as a Se vice (PaaS) e vi o me t, utilizi g mode Co ti uous I teg atio /Co ti uous Delive y (CI/CD) pipeli es.
“We a e ho o ed to co ti ue ou t usted pa t e ship with the IRS o the SimoN i itiative,” said Cha les Che , CEO of Skyma tics. “This Phase 5 awa d is a testame t to the value of AI i p ovidi g the IRS with the ope atio al ce tai ty they eed to i ovate a d mode ize secu ely, completely elimi ati g PII isk f om the testi g e vi o me t.”
“Phase 5 ep ese ts a massive leap fo wa d i scale a d complexity fo the SimoN e gi e,” added B ad Mola de , CTO of Skyma tics. “We a e movi g beyo d basic household agi g to o chest ate complex, multi-yea busi ess a d i dividual tax eve ts, all outputted i p ecise IRS XML schemas. By i jecti g sophisticated f aud typologies di ectly i to the sy thetic data feeds, we a e shifti g e te p ise testi g to the left-empowe i g the IRS to validate system esilie ce a d safegua d taxpaye systems befo e they a e eve exposed to eal-wo ld th eats.”
