# Navigara > Navigara makes engineering teams’ AI agents run at full speed, and proves what that is worth. It connects to a team’s existing development stack (GitHub, GitLab, Jira, Linear) and converts commit-level activity into objective measurements of delivery, quality, AI ROI and agent speed. Most teams start with an AI ROI report, delivered in five days. Navigara is a San Francisco and Prague based startup founded in 2025 by Jirka Bachel (CEO, co-founder and ex-CTO of Onetone.ai, previously built the Seznam.cz browser) and Peter Malina (CTPO, ex-Director of Engineering at Kiwi.com). The company raised a $2.5M seed round in early 2026. Its flagship metric is **Engineering Throughput Value (ETV)** – a per-commit, LLM-graded measure that benchmarks an engineering team’s output against its own pre-AI baseline, instead of relying on activity totals like commit counts or lines of code. ## Core pages - [Home](https://navigara.com/): What Navigara does, who it’s for, the ETV concept in summary. - [About](https://navigara.com/about): Founders, company background, the story of how Navigara was built. - [Pricing](https://navigara.com/pricing): Public pricing. Explore is free (a 14-day cloud trial, up to 1,000 PRs analyzed), Measure from $7 per developer per month for teams up to 30 developers (performance against your pre-AI baseline, reported weekly to quarterly), Pro from $30 per developer per month (adds AI ROI, Agent Speed checks, roadmap alignment, adaptive token limits, Chameleon model routing and CapEx / OpEx reporting), and custom Enterprise with on-prem collector or full on-premises / air-gapped deployment plus guided pilots. - [Prove your AI ROI](https://navigara.com/product/ai-roi): Measure engineering before and after AI, read the gain as capacity you did not have to hire, and check it against the roadmap. Scored from commit history, not surveys or story points; available as a one-off AI ROI report delivered in five days. - [The State of Your AI Transformation](https://navigara.com/product/ai-transformation): A report for boards, investors and technical due diligence. Five questions and twenty-five checks, scored for every team and every repository from the work itself. - [Product Engineering Health](https://navigara.com/product/engineering-health): The framework behind the product. Coding, reviewing and product are three stages of one loop, so speeding one up just moves the constraint to the next; AI adoption is scored per developer on evidence, the whole review pipeline is tracked, every objective is audited for context health, and a nightly agent files a finding when any of it drifts. - [Engineering Performance](https://navigara.com/product/performance): Performance graded per commit and summed per pull request, per developer and per team, all in ETV; average team performance per month is the headline number, benchmarked against the same team twelve months earlier. - [Alignment](https://navigara.com/product/alignment): Measure how much each team contributed to each business priority. - [Token Spend Intelligence](https://navigara.com/product/token-spend): Chameleon routes each task to the cheapest capable model and is trainable per company; token spend is priced against ETV, and a five-dimension adoption profile names what each team should improve next. - [Adaptive Token Limits](https://navigara.com/product/token-limits): The AI budget exception queue, automated. Every request carries delivered ETV, cost per ETV, roadmap alignment and the features/maintenance/tests/docs/fixes split, benchmarked against every other team; limits then move against rules you set, redistributing a fixed envelope instead of raising it. - [CapEx & OpEx Reporting](https://navigara.com/product/capex-opex): Audit-defensible software capitalization, derived from code. - [AI Adoption](https://navigara.com/product/ai-adoption): How every developer really uses AI. Five levels (agent fluency, delegation depth, roadmap alignment, context leverage, agentic autonomy), scored from commits, pull requests and tickets instead of a survey, never averaged, each with the one next step that raises it; the org view shows how many developers are not started, starting, fluent or leading on each. - [Agent Speed](https://navigara.com/product/agent-speed): Shows how much of full speed your agents run at and what holds them back. 25 checks written by top engineering consultants read your tickets, commits and AI spend nightly, and rank each fix by the speed it gives back. - [Security](https://navigara.com/security): Deployment inside your own perimeter: a collector in your network that keeps source code local, a full on-premises or air-gapped install that reaches Navigara not at all, or managed cloud. Also covers the outbound connection list, read-only credentials with end-to-end encrypted Git tokens, SSO/SCIM/audit logs, signed images, retention defaults, and how to revoke Navigara support access. - [Blog](https://navigara.com/blog): Field notes on measuring engineering, ETV, and AI tool impact. - [Press](https://navigara.com/press): All independent media coverage of Navigara. - [Careers](https://navigara.com/careers): Open engineering, product and data roles. - [Contact](https://navigara.com/contact): Talk to the team or book a call to prove your AI ROI. ## Blog - [Is My Engineering Good Enough?](https://navigara.com/blog/is-my-engineering-good-enough): Why CEOs ask whether engineering is fast enough, why it is really a trust problem (not an AI problem), and why a peer benchmark, not AI spend or DORA, is the only thing that resolves it. - [Benchmark Engineering Before You Wire the Money](https://navigara.com/blog/benchmark-engineering-before-you-wire-the-money): For private equity and late-stage VC. Why technical due diligence is expensive guesswork, and how to benchmark a target’s engineering against the companies that returned your funds, before and after close. - [How to Prove ROI on AI Tools in Engineering](https://navigara.com/blog/how-to-prove-roi-ai-tools-engineering): A four-question framework for proving AI tool ROI by measuring roadmap delivery: performance change year over year, work mix, cost per ETV, and roadmap alignment. - [You Can’t Measure ROI on AI Coding Tools If You Never Measured Engineering](https://navigara.com/blog/measure-roi-ai-coding-tools): Why ROI on AI coding tools is unknowable until engineering output itself is measured. - [Performance per Engineer Doubled at Six Big Tech Companies](https://navigara.com/blog/performance-doubled-six-big-tech): Commit-level analysis of 676 engineers showing a 116% year-over-year jump, with a Q1 2026 inflection point. - [The Story of Navigara](https://navigara.com/blog/story-of-navigara): Founder Jirka Bachel’s long-form write-up on why and how the platform was built. ## Customers Navigara’s customers are engineering organizations across travel, fintech, banking, and cybersecurity. Named customers include Kiwi.com (online travel), ESET (cybersecurity), FTMO (proprietary trading / fintech), Finshape (digital banking software), Partners Banka (banking), Purple Technology (fintech), Greyson, Itrinity, kegg, and Second Foundation. ## Reference - [Research](https://research.navigara.com): Public research, data, and the white papers behind ETV. - [Docs](https://docs.navigara.com): Product documentation. ## Optional - [LinkedIn – Jirka Bachel](https://www.linkedin.com/in/jirka-bachel/): Co-Founder & CEO. - [LinkedIn – Peter Malina](https://www.linkedin.com/in/petomalina/): Co-Founder & CTPO. - [llms-full.txt](https://navigara.com/llms-full.txt): Longer machine-readable summary including ETV definition, founder bios, and key facts.