Be li , Ge ma y, Ju e 26, 2026 (GLOBE NEWSWIRE) — Peec AI, the AI sea ch a alytics platfo m, today a ou ced the esea ch team it has assembled to a swe the questio eve y b a d is ow aski g: why do ChatGPT, Pe plexity, a d Google’s AI Mode ecomme d some compa ies ove othe s – a d how ca a b a d ea its place i those a swe s?
Fo me e te p ise SEO leade s, a explai able-AI esea che , a d the e gi ee s behi d some of the ma ket’s ea liest tools a e eve se-e gi ee i g AI sea ch, feedi g what they lea st aight i to a p oduct built to help clie ts ea visibility the e.
As AI a swe e gi es eplace the old list of blue li ks, the ules fo which b a ds get ecomme ded ow sit i side models with o published ulebook – a d o o e outside those compa ies k ows exactly how the choices a e made. Peec AI’s espo se: hi e the people who wo k those ules out by expe ime t, pai i g two decades of e te p ise SEO with ha ds-o eve se-e gi ee i g of a swe e gi es a d machi e-lea i g esea ch i to how models decide.
That esea ch flows st aight i to the p oduct. Peec AI t acks how b a ds appea ac oss the majo a swe e gi es – measu i g thei Ge e ative Sha e of Voice agai st competito s, se time t, a d the sou ces each e gi e cites. Those sig als become a p io itized set of ecomme datio s that show teams how to i c ease thei visibility i AI sea ch e gi es.
Much of that wo k is led by Meteha Yeşilyu t a d Tomek Rudzki. Yeşilyu t eve se-e gi ee s the algo ithmic behavio of AI sea ch a d a swe e gi es without touchi g the u de lyi g models’ sou ce code. I stead, he u s co t olled tech ical expe ime ts, a alyzes API etwo k t affic, a d eads clie t-side co figu atio s to i fe how platfo ms decide which web co te t to su face a d cite.
Yeşilyu t’s fi di gs, which he publishes ope ly o his ow blog, have helped shape the eme gi g fields of Ge e ative E gi e Optimizatio (GEO) a d A swe E gi e Optimizatio (AEO) fo b a ds, i -house teams, a d ma keti g age cies. By a alyzi g ChatGPT’s web i te face a d etwo k t affic, he decoded how the assista t decides which sou ces ea a i li e citatio , su faci g evide ce that the behavio eflects a Recip ocal Ra k Fusio app oach that agg egates a d sco es co te t ac oss multiple subque ies. I late 2025 he published a detailed b eakdow of Pe plexity AI, docume ti g mo e tha 59 disti ct a ki g facto s a d patte s, amo g them the platfo m’s elia ce o a sho tlist of top-tie sou ces, sema tic eleva ce to the que y, co te t f esh ess, a d apid use -e gageme t sig als. He has applied the same method to Google Discove a d Google’s AI Mode, mappi g the multi-stage co te t pipeli es a d f esh ess “buckets” that appea to gove which pages su face.
Tomek Rudzki has bee decodi g ge e ative sea ch e gi es si ce befo e the field eve had a ame. P io to joi i g Peec AI, Rudzki was a co-fou de of ZipTie.dev – the fi st comme cial tool (ea ly 2024) dedicated to mo ito i g b a d visibility i AI sea ch e gi es. Notably, the tool t acked Google AI Ove views f om its ea liest days u de the “SGE” ba e , back whe it was est icted to logged-i use s withi Google Labs. I this ole, Tomek was di ectly espo sible fo desig i g a d buildi g the module that optimized co te t fo la ge la guage models, givi g b a ds a ta gible way to i flue ce whethe a d how AI systems cited thei sou ces. His espo sibilities i cluded buildi g a LLM-a swe se time t a alyze a d a p ompt ge e ato .
That same ha ds-o u de sta di g ow powe s Peec’s AI-visibility ecomme datio s – so a b a d sees ot just how it’s pe fo mi g ac oss AI a swe e gi es, but the specific cha ge most likely to g ow its visibility.
That i sti ct u s deep. Rudzki’s e gi ee i g thesis at Politech ika Opolska was titled Empi ical A alysis of the Google Sea ch E gi e Algo ithm, a d fo his 2019 compute -scie ce maste ‘s he built softwa e that imp oved a ki gs i t aditio al sea ch e gi es usi g SEO c awl data, se ve logs, a d Google Sea ch Co sole sig als. He has spoke at i dust y co fe e ces i cludi g B ighto SEO a d SMX.
That expe ime tal wo k is g ou ded by people who a Google o ga ic sea ch at scale lo g befo e LLMs e te ed the pictu e.
Malte La dweh b i gs mo e tha 20 yea s of e te p ise SEO a d SaaS p oduct leade ship. He p eviously se ved as VP of SEO at idealo, o e of the wo ld’s la gest p ice-compa iso platfo ms, whe e he spe t five yea s leadi g o ga ic g owth a d doubled Google o ga ic t affic, a d as VP of P oduct at Sea chmet ics, a globally ecog ized e te p ise SEO softwa e suite. He left a se io co po ate ole to help build Peec AI – the p oduct he had co cluded the ma ket was missi g.
David Ko itz y b i gs a tech ical-SEO a d i te atio al-sea ch pe spective shaped o both sides of the table – age cy a d i -house. He sta ted out as a tech ical SEO at Peak Ace, o e of Ge ma y’s most e ow ed age cies, befo e becomi g Head of O ga ic Sea ch at Raisi , a Ge ma fi tech u ico , whe e he built multili gual, multi- egio al st ategies ac oss i te atio al ma kets. Whe he etu ed to age cy wo k at KKP, that i -house expe ie ce paid off: he al eady u de stood the pai poi ts faci g i su a ce a d ba ki g clie ts, a d how to solve them. KKP is also whe e AI sea ch became his focus – a d the app oach he developed the e is the o e he b i gs to Peec: ot just optimizi g a clie t’s domai , but educati g the whole ma keti g team a d givi g la ge o ga izatio s a clea oadmap to follow. He pai s that tech ical depth with comme cial i sti ct, a d today uses b owse develope tools a d othe ha ds-o methods to ead the sig als that eveal how LLM systems behave u de the hood.
T a slati g those discove ies i to wo ki g softwa e falls i la ge pa t to D . Melissa Fasol, o e of the co e develope s o the Peec AI platfo m. She ea ed a PhD i computatio al eu oscie ce at the U ive sity of Edi bu gh, specializi g i Explai able AI, sig al p ocessi g, a d machi e lea i g – wo k focused o exactly the p oblem AI sea ch p ese ts: maki g the decisio s of complex, opaque systems t a spa e t a d t ustwo thy.
Befo e Peec AI, she fou ded Tulia AI a sta tup that helps b a ds u de sta d why LLMs ecomme d ce tai compa ies ove othe s – wo k fo which she was sho tlisted fo the I vestec Ea ly Stage E t ep e eu of the Yea Awa d. That app oach alig s closely with Peec’s missio , a d at Peec she helps tu the team’s collective esea ch i to the platfo m’s Actio s module – a ked, oppo tu ity-sco ed ecomme datio s o what a b a d should do ext to imp ove its AI sea ch p ese ce.
“A yo e ca measu e b a d visibility. The ha d pa t is givi g ecomme datio s that actually move the eedle a d ca be scaled i a sustai able way,” said Malte La dweh , Chief P oduct Office of Peec AI. “Ma y tactics that cu e tly wo k well fo sho t-te m gai s i AI visibility will stop wo ki g i a yea . Some eve put you whole SEO visibility at isk! To avoid that, you eed people with deep SEO expe ie ce who a e also able to deco st uct how LLMs wo k.”
Peec AI is a AI sea ch a alytics platfo m that helps ma keti g teams a d age cies mo ito , be chma k, a d imp ove how thei b a ds appea ac oss AI a swe e gi es. With comp ehe sive cove age of LLM platfo ms – i cludi g ChatGPT, Pe plexity, Gemi i, Google AI Ove views a d AI Mode, Claude. Fou ded i Ja ua y 2025 by Ma ius Mei e s, Da iel D abo, a d Tobias Siwo ia, Peec AI is backed by Si gula , 20VC, a d A tle . Lea mo e at peec.ai
Noah Wolffg owth [at] peec.aihttps://peec.ai/
