Cloud SaaS · Individual Engineers · Early Access
Next
Reservoir
Reservoir
The subsurface work engine. It doesn't just answer — it runs the calculation, writes the code, builds the plot, and cites every source down to the page. In the browser. No install.
Citation-first
Inspectable Code
Zero Install
Vendor Neutral
EU-resident
The problem
Every subsurface question needs a calculation behind it. Today that calculation lives in scattered spreadsheets, local scripts and tribal knowledge — disconnected from the source documents that actually justify the numbers.
Engineers with deep domain knowledge spend too many hours on tooling, syntax and file management. Not on interpretation and decisions.
Engineers with deep domain knowledge spend too many hours on tooling, syntax and file management. Not on interpretation and decisions.
What it does
Next Reservoir completes the technical work, not just the answer. It retrieves the right sources from your knowledge base, writes and executes Python to produce the result, cites every claim to its source document and page, and delivers something you can defend in peer review.
Reservoir engineers, petroleum engineers, geologists, petrophysicists. One platform, one knowledge base, shared across a project.
Reservoir engineers, petroleum engineers, geologists, petrophysicists. One platform, one knowledge base, shared across a project.
The interface
Example session · History Match QC
Troll West GOR Study · SCAL QC session · citation pills · inspectable Python · executed in sandbox
Core capabilities
Four principles
01 · Citation-first
Every claim traces
to a source.
to a source.
Inline citation pills show the document, page, and knowledge tier for every factual statement. A structural validation layer runs on every answer — hallucinated citations are stripped automatically before you see them.
02 · Inspectable code
The AI never does
arithmetic in its head.
arithmetic in its head.
All calculations are written and executed as Python in a sandboxed environment. The code is always visible and downloadable. You run the same script in your own environment and get the same number.
03 · Vendor neutral
OPM Flow — open source,
zero licence cost.
zero licence cost.
Next Reservoir generates keyword blocks and Python setups for the simulator you use. You run them in your licensed environment and paste or upload the result back. No vendor lock-in. No preference for one stack over another.
04 · Zero install
Open a browser.
That is it.
That is it.
No software to install, no IT approval, no VPN. The full capability — calculations, code, corpus search — runs in the cloud. Generated code travels to your local environment via the copy-down path whenever needed.
Why this is different
Work engine
Every interaction produces something defensible: a cited source, inspectable code, a reproducible plot, or a draft deliverable. Not a chatbot output. Not something you have to fact-check manually.
Private by design
What you upload stays in your private workspace and is never pooled into a shared corpus. EU-resident inference via AWS Bedrock Frankfurt. Row-level security enforces tenant isolation at the database level and not by convention.
Four knowledge tiers
Admin Library · personal Boxed Library · Space (project) · Session. Every citation badge shows which tier it came from. You always know whether a source is your upload, a project document, or a curated platform reference.
The platform
nextreservoir.com
The subsurface work engine is live. Some users explore already the roadmap, and they have started a conversation about what is really needed. Thanks.
Visit nextreservoir.com →
Follow the build
Progress on LinkedIn.
Technical decisions, workflow experiments and product updates shared openly as the build progresses. Join the conversation.
LinkedIn → Klaas Kostro