clemenss

Lemon AI | Founding Engineer | Onsite in London, UK | Full Time | https://getlemon.ai

Human progress depends on our ability to train the best AI models. Today, compute is no longer the bottleneck — it's data. Lemon AI exists to create synthetic data of such high quality that it is indistinguishable from real data.

* Some of the technical challenges we need help with: improving our dataset explainability capabilities (e.g. multi-dimensional semantic and lexical analysis), building out our multi-agent system to generate the highest quality synthetic data, scaling infra so we can handle petabytes of data

* We work with well-chosen and unique technologies across the fields of software engineering, data science, ML and AI. Some of them are Golang, htmx and late interaction-based semantic comparison

* You bring 2-5 yrs experience as a backend engineer, technical founder or founding engineer. But most importantly, you are exceptionally driven and have an intrinsic interest in solving data scarcity. You are intense but also fun to work with

* Hybrid work model, based in London, with 4 days in the office per week

* Salary range £70-100k plus 0.5-1% equity. Unlimited holiday policy. Fitness perks to support your physical and mental well-being

Additional info https://lemonai.notion.site/Founding-Engineer-118ca0aa3fa680...

To apply, email me at [email protected]

Lemon AI | Founding Backend Engineer | Onsite in London, UK | Full Time | https://getlemon.ai

The biggest bottleneck in AI is no longer compute - it's data. We are building a synthetic data curation solution for LLM training and fine-tuning, making custom LLMs a reality for all.

  - Some of the technical challenges we need help with: improving our dataset explainability capabilities (e.g. multi-dimensional semantic and lexical analysis), building out our multi-agent system to generate the highest quality synthetic data, scaling infra so we can handle petabytes of data

  - We work with well-chosen and unique technologies across the fields of software engineering, data science, ML and AI. Some of them are Golang, htmx and late interaction-based semantic comparison

  - You bring 2-5 yrs experience as a backend engineer, technical founder or founding engineer. But most importantly, you are exceptionally driven and have an intrinsic interest in solving data scarcity. You are intense but also fun to work with

  - Hybrid work model, based in London, with 4 days in the office per week (3 fixed, 1 flexible)

  - Salary range £70-100k plus 0.5-1% equity. Unlimited holiday policy. Fitness perks to support your physical and mental well-being

  - Additional info https://lemonai.notion.site/Founding-Engineer-1879e8ae4db94b8dbff90f9ebdbb788d?pvs=4

To apply, please email me at [email protected]

Lemon AI (https://getlemon.ai) | Founding Engineer | London, UK Onsite | Full Time

* Work on the most exciting problem in the AI space: data scarcity. Lemon AI is building a synthetic data curation solution for LLM training and fine-tuning to make custom LLMs a reality for all

* Backed by Stijn Christiaens (Co-Founder Collibra, valued at $5bn+) & Julien Launay (Co-Founder Adaptive ML, trained Falcon LLM), Index Ventures scout fund & others

* Beautiful office on Baker Street

If you are exceptionally driven, passionate about the emerging AI infra stack and bring 1-2+ yrs experience as a backend engineer, technical founder or founding engineer, please email your resume to clemens at getlemon.ai

More info: https://lemonai.notion.site/Founding-Engineer-1879e8ae4db94b...

Thanks very much for your feedback, much appreciated! I’ll try to fix this

Hi HN! We’re Clemens and Felix from Cito - thrilled to show you what we’ve built to help data engineers stay on top of data quality issues. Think Datadog meets Incident.io.

Tests in dbt are great when checking whether specific expectations are true, but don’t work well for use cases where data patterns may change over time. When relying on testing alone, data teams regularly face situations where business stakeholders identify data issues in dashboards first, eroding trust. In such situations, understanding the implications of an issue and debugging can be a very manual and time-consuming process.

To help data engineers ensure trust in data, Cito makes it easy to go beyond simple tests. By executing scheduled or near real-time out-of-the-box anomaly detection tests (row count, schema change, etc.) or custom SQL tests, data anomalies are detected and communicated in the context of the relevant column-level lineage via Slack.

We believe data observability solutions should not stop at alerting teams to anomalies and our ambition is to support the complete end-to-end workflow of data engineers. Leveraging column-level lineage, our solution makes it straightforward to understand the context of anomalies. In addition, by automatically providing transparency in a git-blame-like fashion around ownership of data models and showing who made changes most recently, Cito helps to accelerate internal communications when troubleshooting.

We’re super keen to hear your thoughts, ideas and experiences! You can also use our docs to try Cito in less than 15 min.

Hi HN! We’re Clemens and Felix from Cito - super excited to launch Cito (https://www.citodata.com) today, the first truly actionable data observability solution built for small data teams.

Many data teams we’ve spoken with spend hours writing and maintaining tests in dbt, only to achieve a poor outcome. Tests are great when testing whether specific expectations are true, but don’t work well for use cases where data patterns may change over time, for example, because of seasonality. When relying on tests only, teams regularly end up in situations where business stakeholders identify silent data issues in dashboards first, eroding trust. In such situations, understanding the implications of an issue and debugging can be a very manual and time-consuming process.

To help teams ensure trust, Cito automatically detects data anomalies and communicates them in the relevant column-lineage context via Slack, allowing users to instantly understand the root cause of an anomaly and its impact. Cito executes tests based on pre-defined frequencies, but it is also possible to execute tests whenever tables refresh in Snowflake, offering a near real-time monitoring experience. Next to our web app, we provide programmatic access via our REST API, e.g. to trigger tests based on external events.

Existing solutions are either expensive ($50-300k/year) or lack actionability. What’s different about Cito is that we leverage column-level lineage and other actionability-enhancing features such as in-app SQL model accessibility to make the incident response super straightforward. This facilitates our users to quickly and easily identify the root cause and downstream impact of anomalies that may occur in their data, allowing them to respond quickly and prioritize appropriately.

We would love to hear your thoughts, ideas and experiences!