Frequently Asked Questions
FAQ contains the latest information about Nettle and supersedes all pre-dated documentation available on the website or dataroom. Last updated: 22nd September, 2026
What is Nettle?
Who is using Nettle today, and in what capacity (paid, pilot, design partner, evaluation, internal demo)? Can any be named or described by industry and size?
What is the design-partner programme: what a partner gets, what they commit to, duration, number of slots, and how many are in discussion at which stage?
What are the current funnel numbers (leads contacted, conversations, demos, trials, pilots in scoping), and as of what date?
What is the single workflow a customer would pay for today if every other feature disappeared?
How does Nettle fit alongside an existing Snowflake, Databricks or dbt stack: replace, sit on top, or run beside? What stays in place on day one?
What is the realistic timeline from first call to value in an enterprise? What takes five minutes, what takes weeks (security review, data access), and what does Nettle do to shorten it?
How will Nettle avoid becoming a bespoke consulting shop for its first partners? What is in and out of scope for partner requests?
In one sentence a data leader would use, what is Nettle, and which existing budget line does it come out of?
What is the current round: size, instrument, use of funds, the milestones it pays for, and timeline?
What can be said about cash position and months of runway, as of a date? What is the plan if the round takes longer than expected?
What is current monthly burn, the board-approved limit, and the hiring plan tied to milestones? Why did go-to-market hires come before design partners?
Is the long-range model in the Q4 2025 report (about $440M raised for a $1.3B exit) still the company's view? If not, what replaces it?
What is the measured distribution of time-to-insight by workload type (rows, tables, source): median and 90th percentile? What exactly does '5 minutes' refer to?
What is the largest dataset processed (rows, tables, size)? Is the system single-node or distributed today? What is tested versus designed for?
How accurate is automated classification and modelling, on what test sets, measured how? When the model is wrong, how does a user find out and correct it?
What can a user review, approve, override and roll back? What is logged, and is there an audit trail?
What escape hatches exist: custom SQL or Python, overriding generated transformations, export, git or dbt integration?
In plain engineering language, what does the formal model change in the implementation: data structures, algorithms, invariants?
Which parts of the product running today implement that model, and which parts are roadmap?
Is there a technical note or paper that can be published? Has anyone outside the company reviewed it?
Who on the full-time team owns the core engine and understands the formal model? What is the current role of the mathematical advisers?
What IP protection exists: patents filed or pending, trade secrets, and invention-assignment agreements covering every contributor including contractors?
What is shipped, in beta, and planned? Provide a simple capability table.
Which LLMs or providers does the product depend on, what data is sent to them, and what happens if a provider changes price or terms?
What is the compliance roadmap (SOC 2, ISO 27001): current status, auditor, target dates?
What customer data is stored, where, for how long, how is it deleted, is it used for training, and which subprocessors see it?
Is everything in the public workspace demo data? Why are IP and location fields visible, and how is the public workspace isolated from customer tenants?
Which deployment options exist today versus planned: SaaS, customer VPC, on-prem, air-gapped?
What are the availability targets, what does the status page cover (production or development), what caused recent incidents, and what are RTO/RPO and backup arrangements?
How does automated relationship discovery treat personal data: PII detection, masking, access policies, GDPR rights?
Has a penetration test been done or scheduled? How are vulnerabilities managed?
If I already run Databricks (Genie Code, Lakeflow), why would I add Nettle?
If I already run Snowflake (CoCo, Dynamic Tables, ZeroOps), why would I add Nettle?
How does Nettle relate to dbt Copilot, Fivetran, Airbyte, Fabric Copilot and Atlan: complement or compete?
What is the current competitive view that replaces the 2025 Porter analysis, including where Nettle loses?
Does Nettle read from and write to Snowflake, Databricks, BigQuery, S3 or Iceberg?
Can customers export their models and data in open formats? What is the lock-in position?
Who is full-time in engineering today, and who is (or when will there be) a technical lead?
Which public links evidence the founder's background: Pinterest Engineering posts, Flink Forward talk, patents?
Can one public dataset be shown start to finish, with timings, generated artefacts and validation results, in a form anyone can replay?
What can an evaluator do in the app in 30 minutes without talking to anyone, and what are the trial limits?
What is the roadmap for the next two quarters, tied to customer and technical milestones?
Will the next round use NVCA-standard documents, and what happens to the current charter?
Frequently Asked Questions
FAQ contains the latest information about Nettle and supersedes all pre-dated documentation available on the website or data room. Last updated: 22nd September, 2026
What is Nettle?
Who is using Nettle today, and in what capacity (paid, pilot, design partner, evaluation, internal demo)? Can any be named or described by industry and size?
What is the design-partner programme: what a partner gets, what they commit to, duration, number of slots, and how many are in discussion at which stage?
What are the current funnel numbers (leads contacted, conversations, demos, trials, pilots in scoping), and as of what date?
What is the single workflow a customer would pay for today if every other feature disappeared?
How does Nettle fit alongside an existing Snowflake, Databricks or dbt stack: replace, sit on top, or run beside? What stays in place on day one?
What is the realistic timeline from first call to value in an enterprise? What takes five minutes, what takes weeks (security review, data access), and what does Nettle do to shorten it?
How will Nettle avoid becoming a bespoke consulting shop for its first partners? What is in and out of scope for partner requests?
In one sentence a data leader would use, what is Nettle, and which existing budget line does it come out of?
What is the current round: size, instrument, use of funds, the milestones it pays for, and timeline?
What can be said about cash position and months of runway, as of a date? What is the plan if the round takes longer than expected?
What is current monthly burn, the board-approved limit, and the hiring plan tied to milestones? Why did go-to-market hires come before design partners?
Is the long-range model in the Q4 2025 report (about $440M raised for a $1.3B exit) still the company's view? If not, what replaces it?
What is the measured distribution of time-to-insight by workload type (rows, tables, source): median and 90th percentile? What exactly does '5 minutes' refer to?
What is the largest dataset processed (rows, tables, size)? Is the system single-node or distributed today? What is tested versus designed for?
How accurate is automated classification and modelling, on what test sets, measured how? When the model is wrong, how does a user find out and correct it?
What can a user review, approve, override and roll back? What is logged, and is there an audit trail?
What escape hatches exist: custom SQL or Python, overriding generated transformations, export, git or dbt integration?
In plain engineering language, what does the formal model change in the implementation: data structures, algorithms, invariants?
Which parts of the product running today implement that model, and which parts are roadmap?
Is there a technical note or paper that can be published? Has anyone outside the company reviewed it?
Who on the full-time team owns the core engine and understands the formal model? What is the current role of the mathematical advisers?
What IP protection exists: patents filed or pending, trade secrets, and invention-assignment agreements covering every contributor including contractors?
What is shipped, in beta, and planned? Provide a simple capability table.
Which LLMs or providers does the product depend on, what data is sent to them, and what happens if a provider changes price or terms?
What is the compliance roadmap (SOC 2, ISO 27001): current status, auditor, target dates?
What customer data is stored, where, for how long, how is it deleted, is it used for training, and which subprocessors see it?
Is everything in the public workspace demo data? Why are IP and location fields visible, and how is the public workspace isolated from customer tenants?
Which deployment options exist today versus planned: SaaS, customer VPC, on-prem, air-gapped?
What are the availability targets, what does the status page cover (production or development), what caused recent incidents, and what are RTO/RPO and backup arrangements?
How does automated relationship discovery treat personal data: PII detection, masking, access policies, GDPR rights?
Has a penetration test been done or scheduled? How are vulnerabilities managed?
If I already run Databricks (Genie Code, Lakeflow), why would I add Nettle?
If I already run Snowflake (CoCo, Dynamic Tables, ZeroOps), why would I add Nettle?
How does Nettle relate to dbt Copilot, Fivetran, Airbyte, Fabric Copilot and Atlan: complement or compete?
What is the current competitive view that replaces the 2025 Porter analysis, including where Nettle loses?
Does Nettle read from and write to Snowflake, Databricks, BigQuery, S3 or Iceberg?
Can customers export their models and data in open formats? What is the lock-in position?
Who is full-time in engineering today, and who is (or when will there be) a technical lead?
Which public links evidence the founder's background: Pinterest Engineering posts, Flink Forward talk, patents?
Can one public dataset be shown start to finish, with timings, generated artefacts and validation results, in a form anyone can replay?
What can an evaluator do in the app in 30 minutes without talking to anyone, and what are the trial limits?
What is the roadmap for the next two quarters, tied to customer and technical milestones?
Will the next round use NVCA-standard documents, and what happens to the current charter?
What is Nettle?
Who is using Nettle today, and in what capacity (paid, pilot, design partner, evaluation, internal demo)? Can any be named or described by industry and size?
What is the design-partner programme: what a partner gets, what they commit to, duration, number of slots, and how many are in discussion at which stage?
What are the current funnel numbers (leads contacted, conversations, demos, trials, pilots in scoping), and as of what date?
What is the single workflow a customer would pay for today if every other feature disappeared?
How does Nettle fit alongside an existing Snowflake, Databricks or dbt stack: replace, sit on top, or run beside? What stays in place on day one?
What is the realistic timeline from first call to value in an enterprise? What takes five minutes, what takes weeks (security review, data access), and what does Nettle do to shorten it?
How will Nettle avoid becoming a bespoke consulting shop for its first partners? What is in and out of scope for partner requests?
In one sentence a data leader would use, what is Nettle, and which existing budget line does it come out of?
What is the current round: size, instrument, use of funds, the milestones it pays for, and timeline?
What can be said about cash position and months of runway, as of a date? What is the plan if the round takes longer than expected?
What is current monthly burn, the board-approved limit, and the hiring plan tied to milestones? Why did go-to-market hires come before design partners?
Is the long-range model in the Q4 2025 report (about $440M raised for a $1.3B exit) still the company's view? If not, what replaces it?
What is the measured distribution of time-to-insight by workload type (rows, tables, source): median and 90th percentile? What exactly does '5 minutes' refer to?
What is the largest dataset processed (rows, tables, size)? Is the system single-node or distributed today? What is tested versus designed for?
How accurate is automated classification and modelling, on what test sets, measured how? When the model is wrong, how does a user find out and correct it?
What can a user review, approve, override and roll back? What is logged, and is there an audit trail?
What escape hatches exist: custom SQL or Python, overriding generated transformations, export, git or dbt integration?
In plain engineering language, what does the formal model change in the implementation: data structures, algorithms, invariants?
Which parts of the product running today implement that model, and which parts are roadmap?
Is there a technical note or paper that can be published? Has anyone outside the company reviewed it?
Who on the full-time team owns the core engine and understands the formal model? What is the current role of the mathematical advisers?
What IP protection exists: patents filed or pending, trade secrets, and invention-assignment agreements covering every contributor including contractors?
What is shipped, in beta, and planned? Provide a simple capability table.
Which LLMs or providers does the product depend on, what data is sent to them, and what happens if a provider changes price or terms?
What is the compliance roadmap (SOC 2, ISO 27001): current status, auditor, target dates?
What customer data is stored, where, for how long, how is it deleted, is it used for training, and which subprocessors see it?
Is everything in the public workspace demo data? Why are IP and location fields visible, and how is the public workspace isolated from customer tenants?
Which deployment options exist today versus planned: SaaS, customer VPC, on-prem, air-gapped?
What are the availability targets, what does the status page cover (production or development), what caused recent incidents, and what are RTO/RPO and backup arrangements?
How does automated relationship discovery treat personal data: PII detection, masking, access policies, GDPR rights?
Has a penetration test been done or scheduled? How are vulnerabilities managed?
If I already run Databricks (Genie Code, Lakeflow), why would I add Nettle?
If I already run Snowflake (CoCo, Dynamic Tables, ZeroOps), why would I add Nettle?
How does Nettle relate to dbt Copilot, Fivetran, Airbyte, Fabric Copilot and Atlan: complement or compete?
What is the current competitive view that replaces the 2025 Porter analysis, including where Nettle loses?
Does Nettle read from and write to Snowflake, Databricks, BigQuery, S3 or Iceberg?
Can customers export their models and data in open formats? What is the lock-in position?
Who is full-time in engineering today, and who is (or when will there be) a technical lead?
Which public links evidence the founder's background: Pinterest Engineering posts, Flink Forward talk, patents?
Can one public dataset be shown start to finish, with timings, generated artefacts and validation results, in a form anyone can replay?
What can an evaluator do in the app in 30 minutes without talking to anyone, and what are the trial limits?
What is the roadmap for the next two quarters, tied to customer and technical milestones?
Will the next round use NVCA-standard documents, and what happens to the current charter?