E gi ee i g p oductivity platfo m Li ea B’s latest esea ch epo ts o codi g boom
Bosto , MA, Nov. 02, 2026 (GLOBE NEWSWIRE) — Li ea B, a Leade i the 2026 Ga t e Magic Quad a t fo Develope P oductivity I sight Platfo ms, today lau ched Age tic Studio, a key pa t of the ext ge e atio of develope p oductivity platfo ms.
Histo ically, p oductivity platfo ms gave e gi ee i g leade s met ics. Li ea B ow tu s eve ythi g happe i g ac oss softwa e delive y i to co text that people a d age ts ca co sume a d act o . E gi ee i g leade s ask questio s i plai la guage a d get a swe s a d ext steps. They defi e the co text su face with ole-based access built i , a d AI assista ts help exte d that co text a d build met ics, a tifacts a d dashboa ds. Codi g age ts eceive that co text, so they u de sta d thei task a d e vi o me t bette a d make sma te decisio s.
I Li ea B’s Age tic Studio, the AI assista t builds a custom met ic f om a plai -la guage equest.
Acti g o that co text matte s mo e tha eve . Li ea B’s 2026 AI be chma ks, cove i g 2.7 millio pull equests ac oss mo e tha 250 o ga izatio s, show that develope s who use AI most heavily ea ly doubled thei me ged pull equests yea ove yea . Yet the softwa e facto y faces ew p oblems. I Li ea B’s 2026 Softwa e E gi ee i g Be chma ks Repo t, AI-assisted pull equests we e 2.6 times la ge tha u assisted o es at the 75th pe ce tile, a d age t pull equests waited mo e tha five times lo ge fo eview.
Li ea B fou de s O i Ke e a d Da Li es.
“Co text is the ew ki g,” said O i Ke e , CEO a d co-fou de of Li ea B. “We ealized we hold a k owledge g aph that obody else has of eve ythi g happe i g ac oss the softwa e delive y lifecycle, a d that’s exactly what age ts a d teams eed to delive high quality softwa e, faste . So we ebuilt Li ea B a ou d it.”
Li ea B’s effo t skill, ow i ea ly access, matches a codi g age t’s effo t to the task. It gauges a cha ge’s busi ess eleva ce, complexity a d isk usi g the code quality i the a ea, ece t se vice i cide ts a d ow e ship histo y. Low isk cha ges use less effo t a d fewe toke s. Complex, isky o high p io ity cha ges get mo e effo t a d ca eful, tested edits. Mo e skills a e o the way, focused o accele ati g dow st eam stages of the softwa e facto y, such as eview a d me ge.
I Li ea B’s Age tic Studio, the AI assista t t aces a cycle time spike to eview delays a d flags age t-ope ed PRs that have waited lo gest.
K ake used Li ea B’s MCP se ve with AI to fi d what was eally slowi g its code eviews. The a swe was ’t la ge cha ges. It was time zo es a d eviewe availability. K ake cut eview time at the 90th pe ce tile by mo e tha half i fou mo ths. “That’s the st uctu al shift,” said Nik Suda , e gi ee i g ope atio s lead at K ake . “It moves leade s f om a ecdote-d ive to evide ce-d ive .”
The Age tic Studio is available ow to all Li ea B custome s. Age t skills a e available i ea ly access.
Why it matte s owE gi ee i g leade s have mo e data tha eve a d less time to use it. Boa ds a d CFOs ow ask whethe AI spe d is payi g off, a d a dashboa d full of umbe s does ’t a swe that. At the same time, codi g age ts w ite mo e of the code, a d they sta t eve y task k owi g othi g about the codebase’s histo y.
P oductivity platfo ms have always bee good at measu i g. The ext step is tu i g what they measu e i to actio . That mea s a swe s fo the people u i g e gi ee i g, a d co text fo the age ts doi g mo e of the wo k.
What Li ea B doesLi ea B collects data f om ac oss the softwa e delive y lifecycle a d makes it usable. It o malizes data f om diffe e t Git a d p oject ma ageme t systems i to o e model, matches each e gi ee ’s ide tities ac oss tools, a d li ks eve y cha ge f om ticket to elease. It adds AI att ibutio a d i dust y be chma ks, so leade s ca tell whethe a umbe is good o bad.
The e gi ee i g co text laye delive s that k owledge to people a d to age ts.
E gi ee i g leade s ask questio s i plai la guage, like whe e delive y is stuck, how thei develope s a e doi g a d whethe AI is payi g off. I stead of eadi g a cha t, they get a a swe a d a ext step. They ca build thei ow met ics a d pull the same co text i to AI tools like Claude th ough MCP.
Codi g age ts get the same co text th ough MCP a d age t skills. The fi st skill, ow i ea ly access, matches a age t’s effo t to the task. It checks how complex a d isky a cha ge is, usi g the code’s ewo k ate, ece t i cide ts a d ow e ship histo y. Simple, low isk cha ges use less effo t a d fewe toke s. Complex o isky cha ges get mo e effo t a d ca eful, tested edits, which aises quality whe e it matte s most. The skill wo ks with a y codi g age t, a d leade s ca see how ofte age ts we e outed to low, medium a d high effo t.
The platfo m is built fo the e te p ise, with ole based access a d gua d ails o what age ts ca do.
Jou ey a d milesto esLi ea B sta ted by givi g e gi ee i g leade s visibility i to how thei teams delive softwa e. Yea s of i teg atio wo k ac oss the softwa e delive y lifecycle gave it o e of the la gest datasets o how e gi ee i g teams wo k. Its 2026 Softwa e E gi ee i g Be chma ks Repo t d ew o mo e tha 8.1 millio pull equests f om 4,800 e gi ee i g teams i 42 cou t ies. As AI codi g tools sp ead, Li ea B added AI att ibutio a d co ected to mo e of the tools teams use.
I 2026, Ga t e amed Li ea B a Leade i its fi st Magic Quad a t fo Develope P oductivity I sight Platfo ms, fo both its ability to execute a d its complete ess of visio . Ea ly custome espo se to the co text laye is st o g. The AI assista t a d MCP access show the highest use ete tio of a y featu es Li ea B has eleased.
“These a e exciti g times to ethi k ou ole i the softwa e ecosystem,” said O i Ke e . “Eve yo e else is focused o w iti g code. We ealized we hold a k owledge g aph of eve ythi g happe i g ac oss the softwa e delive y lifecycle, a d that’s exactly what develope s a d age ts eed. So we ebuilt Li ea B a ou d it. We do ’t build the softwa e facto y. We i st ume t it, so leade s k ow what to do ext a d age ts make sma te decisio s o eve y cha ge.”
“Fo yea s, e gi ee i g leade s told us the same thi g. Met ics a e useful, but they do ’t tell you what to do,” added Da Li es, co-fou de a d COO of Li ea B. “A dashboa d says you cycle time is two days. So what? Now a leade ca ask what to imp ove i thei e gi ee i g o ga izatio a d get a clea a swe . A d whe that same co text goes to codi g age ts, the platfo m is ’t just epo ti g a ymo e. It’s acti g.”
Ma ket t e dsLi ea B’s 2026 AI be chma ks, built f om 2.7 millio pull equests a d 83,000 develope s ac oss 253 e gi ee i g o ga izatio s betwee Feb ua y a d May 2026, show why measu i g AI is ’t e ough o its ow . The teams pulli g ahead a e the o es acti g o what they lea .
- Heavy AI use s a e pulli g away. The typical develope who uses AI o at least 75% of codi g days me ged 96% mo e pull equests pe week tha a yea ea lie . The typical develope usi g o AI me ged 3.6% fewe .
- Adoptio is shallowe tha the hype suggests. Eve at elite o ga izatio s, AI assists o ly 54% of pull equests.
- Age ts still eed people. At the 90th pe ce tile, auto omous age ts ope 4.7% of pull equests. At elite o ga izatio s, 79% of age t pull equests me ge withi 30 days, compa ed with 92% of huma o ly pull equests. At fai o ga izatio s, age t yield d ops to 37%. The likely easo is ow e ship. Whe o develope ow s a age t’s pull equest, it ofte eve me ges.
- AI code eview is the fastest wi . Pull equests with a AI eview me ge at the highest ate of a y type a d lift yield by up to five pe ce tage poi ts.
- AI costs a e climbi g fast at the top. Toke spe d at the 90th pe ce tile eached $35.20 pe develope pe codi g day, up ea ly 75% i th ee mo ths, while the media ba ely moved.
E gi ee i g leade s al eady use Li ea B to a swe the questio s thei boa ds a e aski g about AI. “I use Li ea B data a d cha ts di ectly i my boa d decks, because my boa d is specifically aski g questio s about ou use of AI a d what ou ROI is,” said Ke Ri gdahl, CTO of Embu se.
Fo wa d-looki g visio The effo t skill is the fi st i a lib a y of age t skills. Each o e tu s Li ea B’s co text i to a actio a age t o a leade ca take. Next, Li ea B will focus o helpi g teams me ge a d gove AI ge e ated code safely, a d o cutti g what e gi ee i g o ga izatio s spe d o AI. The goal is the o e p oductivity platfo ms have always p omised, helpi g teams delive bette softwa e faste . Li ea B ow does it by givi g people a swe s a d givi g age ts the co text to act.
Media images ca be fou d he e.
About Li ea BLi ea B is the develope p oductivity platfo m built o a fou datio al e gi ee i g co text laye . It tu s data f om the tools teams al eady use i to co text fo age ts, develope s a d e gi ee i g leade s. Teams use it to fi d bottle ecks, p ove what AI delive s a d guide age ts towa d safe , cost-awa e cha ges. Lea mo e at https://li ea b.io/
CONTACT: Fo fu the i fo matio please co tact Kie ste Gaff ey at Li ea B o kie ste @li ea b.io
