Faros AI Curie Release: AI for Engineering Leaders

In homage to Marie Curie’s transformative discoveries in radioactivity and their application to diagnose medical injuries quickly on the battlefield with mobile X-ray machines, our newest product, the Curie release, diagnoses engineering organizations.

Now, AI provides engineering leaders with the clearest view yet of what's happening in their organization and—more importantly—why and what to do about it.

With the Curie release, Faros AI becomes a copilot for engineering leaders. Leveraging the power of AI over the evergreen, high-quality data that we map, attribute, correlate, and analyze, we’re revealing the hidden reasons for performance issues and recommending solutions.

Here are the new capabilities that will change how you run and optimize your engineering organization:

• Lighthouse AI Insights: Find out what is impacting each team’s performance and receive recommended solutions and mitigation strategies.
• Lighthouse AI Summary: Get the bottom line and key takeaways from every dashboard and chart with natural language summaries.
• A new R&D Cost Capitalization module: Automate this notoriously tedious reporting and benefit from finance-ready reports with a few clicks.
• An improved AI Copilot Evaluation module: A complete framework for capturing the journey from pilot to value.
• Guided Data Wizard: Quickly find and customize existing charts with this step-by-step wizard or design a new query using natural language.
• Dashboard enhancements: Your custom dashboards will look sharper than ever, with new tabs, an improved layout, and auto-hide for empty cards.
• PII Protector: Remove personally identifiable information (PII) from any data before it’s ingested into Faros.

Let’s dive in, starting with Lighthouse AI Insights.

Understanding why with Lighthouse AI Insights

Consider this:

• Team Avengers has longer PR Cycles than average. Useful.
• Team Avengers has longer PR Cycles than average due to cross-geo reviews. Priceless.

Many tech organizations utilize metrics to understand team productivity, business alignment, software quality, and operations effectiveness. But our new Lighthouse AI Insights go far beyond metrics to root cause analysis. We sift through data from tens, hundreds, and thousands of teams to identify what exactly is impacting performance for each one.

Lighthouse AI Insights identifies teams you should look at and explains why they’re doing better or worse than others. It also provides tangible recommendations on how to improve. This eliminates many hours of investigation and allows leaders and their teams to take corrective action much faster.

Here’s an example for a hypothetical “Team Avengers”:

A metric will tell you that Team Avengers’ PR Cycle Time is 70% longer than the average and the longest of all teams.

A Lighthouse AI insight will identify the factors that make Team Avengers different and explain their longer cycle times. For each contributing factor we find, we also provide recommendations on how to solve it.

Now Team Avengers’ engineering manager will have much more confidence that the improvement measures she’s suggesting for the team will indeed deliver the highest impact:

• If code complexity is the main contributing factor, she’ll prioritize refactoring legacy code and handling other tech debt.
• If it’s cross-geo dependencies, she’ll reorg the team to bring repo and initiative ownership into the same geography.
• If it’s too many meetings and interviews, she’ll suggest creating ‘no interruption’ calendar blocks.

Watch this 90-second video to see the insights in action:

FAQs about Lighthouse AI Insights

• Is this simply anomaly detection? While anomalies exist and have the potential to skew any metric, Lighthouse AI goes beyond anomalies to get to the heart of systemic issues impacting the teams with long-term effects.
• How does Lighthouse AI determine which insights are important? For each metric, Lighthouse AI examines the contributing factors for the chosen time window. It analyzes the strength of the relationship and the differences between the teams to generate an impact score.
• How do I use these insights? Every insight is accompanied by suggested actions to help you determine next steps.

A dream come true: TLDR for dashboards and charts

Ethan Mollick, a professor at the Wharton School at the University of Pennsylvania, recently gave his students an assignment to replace themselves at their next job using GenAI. They built amazing tools to automate significant parts of their job so they could focus on higher-value things.

In the Curie release, we’ve also put GenAI to good use to help you understand your organization faster. You’ll now find TLDR summaries for every dashboard and chart.

With Dashboard Summary, Lighthouse AI summarizes the key takeaways from all the charts on your dashboard. You don’t need to examine each chart separately to understand what your dashboard is saying.

Dashboard Summary creates the TLDR summary from any engineering metrics dashboard

For individual charts, Chart Summary delivers the bottom line from the chart and Chart Explainer describes the purpose of the chart and how it works.

Chart Summary helps engineering leaders interpret the data and Chart Explainer explains how a chart works

Automated R&D cost capitalization reporting

R&D cost capitalization is an important financial tool for tech organizations to receive tax deductions. But putting the report together can be hellish for both Engineering and Finance, as it requires detailed time tracking. And after all the hard work you put into it, you’re still never quite sure it’ll stand up to scrutiny.

That all changes with the new R&D Cost Capitalization module from Faros AI. Gone are the spreadsheets and frantic nights fretting over this task. Now you can get finance-ready auditable reports for monthly, quarterly and annual reporting and continuous views into R&D capitalizable work.

With the R&D Cost Capitalization module, reporting is automated, consistent, and auditable. Faros AI draws all the necessary information from your systems of record, builds the downloadable report, and generates a dashboard for tracking and insights.

Getting started is easy. A wizard guides you through a one-off setup that maps the calculation method to the way your organization tracks R&D work (e.g., by initiative or epic) and translates effort to time (e.g., by story points, time in-progress or other).

Watch this 1-minute video to see how it works:

What is a module in Faros AI? Modules are prebuilt analytics libraries — inclusive of all the data sources, metrics, dashboards, widgets, and customizations you need — that run on top of the Faros AI platform. Infused with domain expertise, benchmarks, and best practices, modules provide rapid insight immediately upon connecting to your data sources. From there, you can build upon the module’s foundation by creating your own custom metrics, views, and reports.

Maximizing the value of coding assistants with an improved AI Copilot Evaluation module

Last October, we released our AI Transformation module to help organizations navigate the adoption of new AI coding assistants like GitHub Copilot and Amazon CodeWhisperer. The module provides a view into time savings, developer sentiment, and downstream impacts to help organizations:

• Track adoption and use over time
• Measure the time savings and economic benefit
• Monitor speed, quality, and security to mitigate unintended consequences

Since the initial launch, we’ve partnered with many enterprises to assist with their adoption, and we’ve infused everything we learned back into the module.

We have renamed the module AI Copilot Evaluation because AI coding assistants are the first frontier of the AI-augmented developer and their adoption deserves a dedicated framework to analyze and maximize impact.

Key additions include:

• A coding assistant value journey, from initial roll-out to larger scale deployments and long-term value optimization
• Granular adoption and usage metrics, including breakdowns by organization, coding languages, and editors
• Out-of-the-box developer surveys to quantify time savings and gain insights into benefits and potential issues
• Benchmarks on the short-term impact you can expect for key velocity and quality metrics such as PR Merge Rate, Review Time, Test Coverage, and PR Size
• Emerging bottleneck identification to help you unlock larger and longer-term impact

Take a 4-minute tour of the AI Copilot Evaluation experience on Faros AI:

Data self-service made easier with guided exploration

Self-service has long been a valued capability for any shared engineering service. Data access should be no different. The guided data exploration wizard helps new Faros AI users get comfortable leveraging data to do their jobs better.

Don’t know where to start? Click Guide Me and use this step-by-step wizard to:

• Choose from a list of common questions people ask of their engineering data.
• Learn to customize existing charts by easily changing filters, groupings, and visualizations.
• Build new charts from scratch using a natural language prompt with Lighthouse AI Query Helper.

After using guided data exploration a few times, users will become familiar with the wealth of charts within Faros AI, the data model, and the tool’s terminology.

Dashboard goodies galore

Curie packages up all the goodness from recent Metabase releases. Metabase provides the front-end BI layer that allows users to view, modify and create their own custom dashboards and charts.

Here are a few highlights we think Faros AI users will adore:

• Keep dashboards organized with tabs. Tabbed navigation helps consume the dashboard in smaller logical bites while persisting filters from tab to tab.
• Sharing is easier with downloadable dashboards. Skip the screenshots! One-click PDF export is now available for dashboards.
• Improved dashboard layout with an expanded grid. Get the polish and symmetry you like with a dashboard grid that’s grown from 18 to 24 squares.
• Auto-hide cards with no data. No data equals no clutter with this new feature. Keep things tidy and easier to consume by setting cards to auto-hide if they return no results.
• Load dashboards faster. You control when to apply filters instead of auto-applying them upon each change.
• Auto-wire up dashboard field filters. Filters are automatically connected to new cards you add to a dashboard.

Introducing PII Protector for PII scanning and redaction

Faros AI is a solution built for enterprises, so we invest heavily in security features. In a world rampant with cyber-attacks and data breaches, removing PII & sensitive data should be essential for any modern data platform.

The ability to redact PII from any data pulled into Faros AI makes your security team happy for its multiple benefits:

• Eases legal compliance with standards like GDPR and CCPA
• Reduces the impact of data breaches
• Serves to deter potential hackers

And as the internal champions of Faros AI, PII scanning and redaction gives you the peace of mind to share data and dashboards broadly to your users without worrying about leaking sensitive information or constantly monitoring your data.

That’s why, in the Curie release, we’ve enhanced our capabilities to remove personally identifiable information and sensitive data with PII Protector. While our customers were always able to prevent specific fields from being pulled into Faros, now they can also redact sensitive data from fields Faros AI is allowed to pull.

How does it work?

• Faros AI provides built-in support for redacting common PIIs such as names, phone numbers, addresses, credit card numbers, social security numbers and more.
• In addition, customers can extend these common fields with custom redaction patterns.

PII Protector is available as part of the Enterprise Bundle.

- - -

Well, that's all folks! And, admittedly, it's a lot. We can't wait to bring you more exciting capabilities next quarter. To learn more about these capabilities or speak to Sales, reach out to our team.