Measure AI’s real impact on engineering[Skip to content](#main) [Pensero](/) [Book a scoping session](/book-demo) # AI is increasing output, not necessarily performance. Pensero turns AI activity into measurable impact by connecting cost to delivery, quality and efficiency, so leaders can make confident decisions. - No surveys - No surveillance - From work you already do [Book a scoping session](/book-demo) Pensero Signals AI adoption AI intelligence AI-assisted Agents Engineering intelligence Work Quality Efficiency Reviews Company Contributors Benchmark Calibrate Settings Org settings Integrations Data health Help center MDMaria Diaz AI intelligence 512 Eng. + AgentsJul 28 – Oct 25 AI economicsContributorsAI budget Efficiency38% $1.42 $/point AI cost21% $255.6k in this period Compliance8pp 92% 471 of 512 Excess spend over budget7% $31.6k over the plan Delivery mix How work spreads across maturity levels 1-5: human-authored, AI-assisted and agents How much delivery lands at each agentic maturity level? 6,010 pts · 67% at level 4+ · 33% below JanMarAprMayJunJulAugSepOct - Human-authored - AI-assisted - AI-augmented review - Orchestrated agents - Autonomous agents How is delivery progressing across the agentic maturity levels? You're at level 4 of 5 · Orchestrated agents — Next: let agents ship low-risk PRs without a human (level 5) JanMarAprMayJunJulAugSepOct - Human-authored - AI-assisted - AI-augmented review - Orchestrated agents - Autonomous agents Efficiency & cost How efficiently we're using AI and what it costs How much does it cost to deliver a point? $1.42 per point JanMarAprMayJunJulAugSepOct What is our AI model usage costing? $5,870 in this period JanMarAprMayJunJulAugSepOct - AI-assisted cost - Agentic cost $/point distribution KTLO$2.86 Dev Productivity$2.12 Improvement$1.58 New Stuff$1.21 Other$0.94 Daily AI cost MonWedFri OctNovDecJanFebMarAprMayJunJulAugSep - No usage - Included in plan - Low - Moderate - High - Very high Adoption Who's using AI coding tools and who's using agents How many active devs are AI-assisted? 79.8% of active devs JanMarAprMayJunJulAugSepOct - AI-assisted devs - Agentic only - Not AI-assisted How many active developers are using agents? 46% of active devs JanMarAprMayJunJulAugSepOct - Agentic users - Non-agentic users Who's getting value? Distribution of AI efficiency across developers Cost per delivery point Cost efficiency ($/pt) Low delivery, High efficiency High delivery, High efficiency Low delivery, Low efficiency High delivery, Low efficiency MDJWLOOHSRDKEMLSPNNBAKTN Delivery ![Pensero AI intelligence page with example data: AI cost per point, AI cost, compliance and spend over budget, then weekly delivery by agentic maturity level, from human-authored work to autonomous agents, the cost per point, the daily AI cost, AI and agent adoption, and the cost efficiency of each person.](/assets/ai-intelligence-bottom.B0bvok0d_10kwoR.webp) ## Example reading. One customer workspace, last 90 days. AI assisted : 39% : of merged code : 8pp adoption Delivery lift : 1.2x : more output shipped : 22% Quality tax : 45.1% : of PRs reworked : 13.2pp Efficiency : 7,891 : tokens per point : 34% worse Extra AI cost : $350K : projected this year : 4.6x ## Engineering teams already measuring with Pensero - ![Perk](/assets/perk.DIhhlz_V.svg) - [![ClosedLoop](/assets/closedloop.LndNNhWa.svg)](/customers/closedloop) - [![Stord](/assets/stord.BXJiX-0z.svg)](/customers/stord) - ![Tuio](/assets/tuio.DIcICsWC.svg) - [![Persuit](/assets/persuit.qooajJ_O.svg)](/customers/persuit) - [![SEAT CODE](/assets/seat-code.Y7Tj-94e.svg)](/customers/seatcode) - [![Cubbo](/assets/cubbo.DrZVo-Jj.svg)](/customers/cubbo) - ![Elfie](/assets/elfie.CMU2D-DP.svg) - ![Caravelo](/assets/caravelo.Bqd9IFLJ.svg) - ![ShopMy](/assets/shopmy.CzXp_cHg.svg) - ![iAhorro](/assets/iahorro.1UZLRwjC.svg) - ![Hum Capital](/assets/hum.DEUDy0_m.svg) - [![Exoticca](/assets/exoticca.PyF3BVAE.svg)](/customers/exoticca) - [![Zynap](/assets/zynap.Bwa6Rd5C.svg)](/customers/zynap) - ![Perk](/assets/perk.DIhhlz_V.svg) - [![ClosedLoop](/assets/closedloop.LndNNhWa.svg)](/customers/closedloop) - [![Stord](/assets/stord.BXJiX-0z.svg)](/customers/stord) - ![Tuio](/assets/tuio.DIcICsWC.svg) - [![Persuit](/assets/persuit.qooajJ_O.svg)](/customers/persuit) - [![SEAT CODE](/assets/seat-code.Y7Tj-94e.svg)](/customers/seatcode) - [![Cubbo](/assets/cubbo.DrZVo-Jj.svg)](/customers/cubbo) - ![Elfie](/assets/elfie.CMU2D-DP.svg) - ![Caravelo](/assets/caravelo.Bqd9IFLJ.svg) - ![ShopMy](/assets/shopmy.CzXp_cHg.svg) - ![iAhorro](/assets/iahorro.1UZLRwjC.svg) - ![Hum Capital](/assets/hum.DEUDy0_m.svg) - [![Exoticca](/assets/exoticca.PyF3BVAE.svg)](/customers/exoticca) - [![Zynap](/assets/zynap.Bwa6Rd5C.svg)](/customers/zynap) - ![Perk](/assets/perk.DIhhlz_V.svg) - [![ClosedLoop](/assets/closedloop.LndNNhWa.svg)](/customers/closedloop) - [![Stord](/assets/stord.BXJiX-0z.svg)](/customers/stord) - ![Tuio](/assets/tuio.DIcICsWC.svg) - [![Persuit](/assets/persuit.qooajJ_O.svg)](/customers/persuit) - [![SEAT CODE](/assets/seat-code.Y7Tj-94e.svg)](/customers/seatcode) - [![Cubbo](/assets/cubbo.DrZVo-Jj.svg)](/customers/cubbo) - ![Elfie](/assets/elfie.CMU2D-DP.svg) - ![Caravelo](/assets/caravelo.Bqd9IFLJ.svg) - ![ShopMy](/assets/shopmy.CzXp_cHg.svg) - ![iAhorro](/assets/iahorro.1UZLRwjC.svg) - ![Hum Capital](/assets/hum.DEUDy0_m.svg) - [![Exoticca](/assets/exoticca.PyF3BVAE.svg)](/customers/exoticca) - [![Zynap](/assets/zynap.Bwa6Rd5C.svg)](/customers/zynap) - ![Perk](/assets/perk.DIhhlz_V.svg) - [![ClosedLoop](/assets/closedloop.LndNNhWa.svg)](/customers/closedloop) - [![Stord](/assets/stord.BXJiX-0z.svg)](/customers/stord) - ![Tuio](/assets/tuio.DIcICsWC.svg) - [![Persuit](/assets/persuit.qooajJ_O.svg)](/customers/persuit) - [![SEAT CODE](/assets/seat-code.Y7Tj-94e.svg)](/customers/seatcode) - [![Cubbo](/assets/cubbo.DrZVo-Jj.svg)](/customers/cubbo) - ![Elfie](/assets/elfie.CMU2D-DP.svg) - ![Caravelo](/assets/caravelo.Bqd9IFLJ.svg) - ![ShopMy](/assets/shopmy.CzXp_cHg.svg) - ![iAhorro](/assets/iahorro.1UZLRwjC.svg) - ![Hum Capital](/assets/hum.DEUDy0_m.svg) - [![Exoticca](/assets/exoticca.PyF3BVAE.svg)](/customers/exoticca) - [![Zynap](/assets/zynap.Bwa6Rd5C.svg)](/customers/zynap) Why Pensero is different ## Is AI making your business better in a way you can prove? Activity · what most teams measure ### Visibility into usage, not impact. - Self-reported surveys - Generic industry reports - Vanity metrics ( PRs, deploys, activity) Counts effort. Says nothing about whether it paid off. Performance · what Pensero measures ### A clear, objective view of ROI. - Real production data - Complexity-weighted delivery - Actual engineering outcomes Ties cost to delivery, quality and efficiency. The questions leaders actually ask. The product ## How Pensero measures AI impact 1. 01 Adoption ### See how much AIyour team really uses Adoption is more than seat counts. Pensero measures how much merged code is AI assisted and how many developers genuinely rely on AI tools. We measure - Share of merged code that is AI assisted - Active developers using AI tools - Adoption trend, month over month - Your numbers against global benchmarks No surveys. Adoption is read directly from the work, then compared to real peers. How many active devs are AI-assisted? 79.8% of active devs JanMarAprMayJunJulAugSepOct - AI-assisted devs - Agentic only - Not AI-assisted W43Oct 19 – 25 79.8%AI-assisted devs 6.0%Agentic only ![Pensero AI adoption chart with example data: 79.8% of active developers are AI-assisted this week, up from 56% in January.](/assets/ai-adoption.COZw0rvR_Z2fDHHe.webp) 2. 02 Impact on delivery & quality ### Connect AI usageto what you ship More output is not the same as better output. Pensero shows whether AI is lifting delivery, and what it costs you in rework. We connect - PR delivery trend as adoption rises - Delivery lift, isolated from headcount - Quality tax: the share of PRs reworked - Out-of-office adjusted, so trends are real The same view that shows a 1.2× lift also surfaces the rework it created. How much delivery is AI-assisted? 1,773 AI-assisted PR points, 30% of PR delivery JanMarAprMayJunJulAugSepOct - Human-only - AI-assisted - AI-augmented review - Agentic W43Oct 19 – 25 4,050.7Agentic 1,370.3AI-augmented review 402.7AI-assisted 186.3Human-only ![Pensero AI delivery chart with example data: weekly PR delivery points split into human-only, AI-assisted, AI-augmented review and agentic work. Agentic work grows from a small share in January to most of the delivery in October.](/assets/ai-delivery.DydIAqm5_Z1olUhc.webp) 3. 03 Efficiency & cost ### Tokens per pointfuel economy for AI Like litres per 100 km: the more tokens it burns to ship one delivery point, the worse the mileage. Right now it is climbing. We track - Tokens per delivery point, over time - Daily AI spend, one square per day - Projected annual extra cost - Spend mapped to output, not raw usage Cost is tied to delivery, so you can see efficiency degrade before the bill does. How much does it cost to deliver a point?$/point$/1M tokenstokens/point $1.42 per point JanMarAprMayJunJulAugSepOct - $/point W43Oct 19 – 25 $1.50$/point 79.8%AI-assisted devs ![Pensero AI efficiency chart with example data: AI costs $1.42 per delivery point this period, down from $2.30 in January, while 79.8% of active developers are AI-assisted.](/assets/ai-efficiency.B3kOd2Gn_Z1JDY7u.webp) 4. 04 Who's getting value ### Spot your high-leveragedevelopers Tokens per delivery point, mapped across your team. We surface - Efficiency distribution across developers - High-leverage developers, by name - Coaching opportunities, made visible - Delivery × efficiency, in four quadrants Distributions are framed for coaching and enablement, not ranking individuals against each other. ![Pensero chart with example data: one dot per developer, placed by delivery and AI efficiency, in four quadrants from low delivery and low efficiency to high delivery and high efficiency.](/assets/value-quadrant.64YcJ2hL_ZysPix.webp) 5. 05 Usage breakdown ### Know which tools andmodels drive delivery Pensero maps every assistant and model in your stack to the delivery it actually produces, so tool decisions are grounded in impact. We break down - Tool mix over time, by share of AI delivery - Model mix across every assistant - Where adoption is concentrating - Trends as your stack evolves Renegotiate, consolidate, or double down, with evidence instead of vendor decks. ![Pensero chart with example data: which AI coding tools are used, week by week from July to June. The most used is Claude Code, at 68.0%, then Cursor, GitHub Copilot, Gemini Code Assist, OpenAI Codex, Cline and Asterix.](/assets/tool-mix.CwovQ5t7_Z1tJBds.webp) Why Pensero ## Built for teams that can't compromise on trust - ### Built on real engineering work Signals from code, tickets and planning tools, translated into execution insight, not self-reported activity. - ### Designed for scaled SaaS Created by former CTOs who have built and scaled complex software organizations through real growth. - ### Aligned across functions A shared execution language for engineering, finance and leadership, grounded in global talent benchmarks. At a glance ## Proven at scale of datapoints processed : Billions developers on the platform : 5k+ onboarding to first value : 15′ increase in performance : 30% ### Trust and compliance, by design Designed to meet strict SOC 2 Type II standards, follows GDPR principles to protect personal data, and supports HIPAA requirements for sensitive information. - ![Secured by Sprinto: SOC 2](/assets/soc2.DAaoUSso_1pbddo.webp) - ![Secured by Sprinto: GDPR](/assets/gdpr.DjuMXYho_Z1OplEv.webp) - ![Secured by Sprinto: HIPAA](/assets/hipaa.CC8NS4_H_1jLqYd.webp) - ![SEAT CODE](/assets/seat-code.Y7Tj-94e.svg) > “We need a way that is objective to see that we’re increasing productivity.” Jordi AbadCTO, SEAT CODE[Read story →](/customers/seatcode) - ![ClosedLoop](/assets/closedloop.LndNNhWa.svg) > “With Pensero we go 4x faster.” Andrew EyeCEO and Founder, ClosedLoop[Read story →](/customers/closedloop) - ![Exoticca](/assets/exoticca.PyF3BVAE.svg) > “Until Pensero, that visibility did not exist.” Raul GómezVP of Technology, Exoticca[Read story →](/customers/exoticca) - ![Cubbo](/assets/cubbo.DrZVo-Jj.svg) > “Now we went from 70% to 30% maintenance… it’s where we want it to be.” Ignasi VegasCEO and Co-founder, Cubbo[Read story →](/customers/cubbo) - ![Zynap](/assets/zynap.Bwa6Rd5C.svg) > “I go every morning into the email, check the first KPIs and drill down.” Jordi MiróCTO, Zynap[Read story →](/customers/zynap) - ![SEAT CODE](/assets/seat-code.Y7Tj-94e.svg) > “We need a way that is objective to see that we’re increasing productivity.” Jordi AbadCTO, SEAT CODE[Read story →](/customers/seatcode) - ![ClosedLoop](/assets/closedloop.LndNNhWa.svg) > “With Pensero we go 4x faster.” Andrew EyeCEO and Founder, ClosedLoop[Read story →](/customers/closedloop) - ![Exoticca](/assets/exoticca.PyF3BVAE.svg) > “Until Pensero, that visibility did not exist.” Raul GómezVP of Technology, Exoticca[Read story →](/customers/exoticca) - ![Cubbo](/assets/cubbo.DrZVo-Jj.svg) > “Now we went from 70% to 30% maintenance… it’s where we want it to be.” Ignasi VegasCEO and Co-founder, Cubbo[Read story →](/customers/cubbo) - ![Zynap](/assets/zynap.Bwa6Rd5C.svg) > “I go every morning into the email, check the first KPIs and drill down.” Jordi MiróCTO, Zynap[Read story →](/customers/zynap) - ![SEAT CODE](/assets/seat-code.Y7Tj-94e.svg) > “We need a way that is objective to see that we’re increasing productivity.” Jordi AbadCTO, SEAT CODE[Read story →](/customers/seatcode) - ![ClosedLoop](/assets/closedloop.LndNNhWa.svg) > “With Pensero we go 4x faster.” Andrew EyeCEO and Founder, ClosedLoop[Read story →](/customers/closedloop) - ![Exoticca](/assets/exoticca.PyF3BVAE.svg) > “Until Pensero, that visibility did not exist.” Raul GómezVP of Technology, Exoticca[Read story →](/customers/exoticca) - ![Cubbo](/assets/cubbo.DrZVo-Jj.svg) > “Now we went from 70% to 30% maintenance… it’s where we want it to be.” Ignasi VegasCEO and Co-founder, Cubbo[Read story →](/customers/cubbo) - ![Zynap](/assets/zynap.Bwa6Rd5C.svg) > “I go every morning into the email, check the first KPIs and drill down.” Jordi MiróCTO, Zynap[Read story →](/customers/zynap) How it works ## Get started in minutes. See insights in hours. Connect your stack. See months of history from day one. 1. 1 hour ### Connect & Structure Integrates repositories, tickets and messages automatically so AI can break work into measurable units, giving your team a unified, transparent view of everything delivered. 2. 1 day ### Analyze & Detect Surface delivery vs roadmap gaps, AI contribution patterns, and execution insights so bottlenecks and early performance drag become visible through evidence-based reviews. 3. 1 week ### Report & Decide Establish a single source of execution truth with team-level and AI vs human performance comparisons, enabling board-ready reporting and confident, scalable capital decisions. Next step ## Every program starts with a baseline In 30 minutes, we scope your audit with you: the teams, the tools and the numbers to measure. [Book a scoping session](/book-demo)[Start free](/auth/signup/) © 2026 Product shots use illustrative data. Platform - [Next moves](/platform/next-moves) - [Agentic ROI](/platform/agentic-roi) - [AI Deployment](/platform/ai-intelligence) - [Delivery Intelligence](/platform/delivery-intelligence) - [Benchmarking](/platform/talent-benchmarking) - [Code Quality](/platform/code-quality) - [CapEx & Finance](/platform/capex-analysis) - [Pensero MCP](/platform/mcp)New - [Integrations](/integrations) Use cases - [CEOs](/use-cases/roles-ceo) - [CTOs](/use-cases/roles-cto) - [CFOs](/use-cases/roles-cfo) - [Product leaders](/use-cases/roles-product-leaders) - [Engineering managers](/use-cases/roles-engineering-managers) - [Investors](/use-cases/roles-investors) - [Compare Pensero](/use-cases/competitors-dx) Company - [About](/about) - [How it works](/how-it-works) - [Careers](/careers) - [Blog](/blog) - [LinkedIn](https://www.linkedin.com/company/penseroai/) - [Support](/support) - [Other resources](/blog/other-resources) - [Glossary](/glossary) Offices - [Barcelona](https://maps.app.goo.gl/6pfxKiKFNKzLPPnB6) - [Madrid](https://maps.app.goo.gl/jEpYBUR4E6AqZs4R6) - [New York](https://maps.app.goo.gl/szk58Jf2MBq14MgAA) - San FranciscoSoon Legal - [Privacy policy](/privacy-policy) - [Cookie policy](/cookie-policy) - [Terms of service](/terms) - [DPA](/dpa) - [Security](https://pensero.trust.site)