Founding Engineer

About Inlet

AI timekeeping for legal professionals.

Tech

We're an applied AI company. Much of what we do is building LLM pipelines that turn messy work artifacts - calendars, emails, documents, browser activity - into clean, defensible time entries that lawyers can actually put on an invoice. Some of the interesting problems we're solving: - **Inference under ambiguity.** In real-world environments, billable work can be difficult to classify. We have to reconstruct what an attorney did, for which client, for how long, from sparse and noisy signals. - **Trust and explainability.** An attorney won't submit a time entry they can't defend to a client or a judge. Every suggestion we make has to be inspectable, editable, and traceable back to the underlying evidence. - **The long tail of legal practice.** Billing conventions vary by practice area, firm, and even across individuals. We can't ship one rigid system - it has to adapt to how each firm actually works. **Stack:** Python / React, Postgres, & GCP. We lean heavily on Claude Code and Codex in our engineering workflow, and we've built a lot of internal tooling to run our go-to-market. We're a small team and the surface area is wide - engineers ship across the stack, talk directly to customers, and own problems end to end.

The role

## **About Inlet**Inlet is building automatic time-tracking software for legal professionals.Lawyers and paralegals track their work in six-minute increments. That means either constantly stopping to record work or reconstructing it later from emails, documents, and memory. It is tedious and impossible to do perfectly. When billable work slips through the cracks, a lawyer can lose thousands of dollars in revenue every month.Inlet captures that work automatically. We help legal professionals spend less time creating time entries and recover more of the work they have already done. And it works: our revenue is doubling every few months and we have strong product-market fit.## **The role**You will be our founding engineer. You will report directly to our CTO and work closely with both founders on what we build, how we build it, and where the product goes next.Automatic timekeeping is a challenging thing to get right. We receive streams of various types of user activity and must determine what work happened, which legal matter it belongs to, how it should be described, and how much time should be recorded. We must do this reliably and as fast as possible, reasoning over noisy signals and producing results legal professionals can trust.The kinds of problems you will help us tackle:* **Make our processing faster.** Help architect and optimize the low-latency systems that turn activity streams into time entries.* **Improve our modeling.** Build evaluation pipelines, identify where accuracy breaks down, run experiments, and make improvements across our multi-stage modeling pipeline.* **Make reviewing time effortless.** Talk with attorneys, paralegals, and customer success, then improve the product and UX so users can review their time with fewer clicks and corrections.* **Build out our development pipeline.** Help us automate parts of our engineering process so we can continue to speed up how we ship.## **Who you are**You have built real software. You can investigate problems without a playbook and make good tradeoffs when speed, quality, and complexity pull in different directions.You love building great products. You want to understand why a customer is struggling before deciding what to build, and you will cross boundaries between backend systems, models, data, and interface design to get the result right.You learn quickly and take ownership. When something you built reaches production, you still consider it yours.Prior experience in legal technology is not required. Neither is prior experience with our exact stack. We want someone who can get deep in unfamiliar territory and emerge with a working, well-reasoned solution.## **Why join Inlet**You will be joining after the product has demonstrated real demand but early enough to shape how it is built. The feedback loop is clear: make Inlet more enjoyable to use and more legal professionals will love and use it!You will have direct access to technical founders, meaningful influence on architecture and product decisions, and the chance to define the engineering culture rather than inherit it.If it sounds fun to build a challenging product that automates something people hate doing, we would love to hear from you!