Bilanc: Engineering Intelligence for the AI Era

The question changes faster than the dashboard

An engineering leader wants to know why reviews slowed down. A week later, the question is whether a new coding tool changed delivery or simply increased spend.

The answers sit across pull requests, tickets, reviews and tool usage. Pulling them into a report can become another request for someone on the team.

Bilanc analyses engineering work and gives leaders a way to ask new questions of that data. Its AI agent creates custom dashboards, reports, analyses and internal tools. [1][2]

The same engineering data can support a recurring report and a question nobody anticipated when the dashboard was built. [2][3]

Start with the work behind the numbers

Bilanc reads pull-request diffs and classifies the work. It excludes generated code and lock files, then assigns a complexity score with a written explanation. [3]

The platform separates cycle time into coding, pickup and review stages. Leaders can see whether work is waiting for attention or taking longer to implement. [3]

Bilanc’s copilot answers questions with links back to the pull requests behind the answer. Scheduled reports bring the analysis into Slack and email. [3]

Integrations bring version control, project management and AI coding-tool data into one model. Teams can use the tools they already work in while Bilanc builds the analysis around that activity. [1]

An agent that builds the view you need

Engineering leaders use the new Bilanc AI agent to build tools around their own engineering-management questions. The request determines the view they create. [2]

One example is an AI-usage view that can be shared across a team. Another is analysis of repository Codeowners settings to help improve review throughput. [2]

Those requests start with a management problem. The agent turns the available engineering data into a dashboard, report or internal tool that addresses it. [2]

Bilanc also offers Data Studio for natural-language queries over its warehouse and an API for custom tooling. An analysis can grow beyond the views shipped with the product. [3]

News UK: delivery and cost in the same discussion

News UK’s technology leaders previously asked delivery leads and principal engineers to assemble data manually. Investment discussions depended heavily on subjective assessments. [4]

Bilanc brings together GitHub and Jira activity with the size and cost of teams. Leaders can compare the output delivered with the resources behind it. [4]

The platform also helped News UK investigate an outlier in AI inference consumption. The follow-up identified gaps in how the team understood the tool and gave leaders a specific coaching issue to address. [4]

For News UK, the change was having the data available when an investment question arose, without starting another round of manual requests. [4]

Engineering data in the agents teams already use

Akinwunmi founded Bilanc in 2023. The company joined Y Combinator’s Winter 2024 batch, building for CTOs, engineering managers and other leaders responsible for delivery. [5]

Bilanc combines engineering analytics, pull-request summaries, release notes and issue refinement. Its new agent extends that work into tools tailored to a customer’s questions. [2][5]

Bilanc’s MCP server also makes engineering analytics and code available to compatible clients such as Claude and Cursor. Access is read-only and follows the caller’s permissions through row-level security. [6]

The direction is to let a leader investigate the work, create the right view and share the resulting analysis in the place the team already uses. [2][3][6]

The bigger picture

Engineering teams need answers that reflect their own repositories, review practices and use of AI.

Bilanc combines a model of that work with an agent that can build around a new question. A fixed report becomes a starting point for investigation. [2][3]

That is engineering intelligence for a team whose tools and questions keep changing.