Product & Research Pipeline
We don't call it done until we prove it works.
On this page you can find the tracker of our research and software portfolio. We take an evidence-based approach to product maturity — that means we have clear passing criteria to determine the stage of our product offerings. It affects both how we structure engagements and our pricing strategy. See detailed explanation below.
How to read the pipeline
Data GovernanceKnowledge Engineering
Data IntegrityAI Safety
Portfolio & Project Mgmt
Software & Dev ToolsWorkflow Orchestration
Research AssistanceOpen Science
Formal MethodsStochastic Process ModelingCreative Problem Solving
Product and research pipeline
Every product, open-source tool and paper, the stage it has reached and what that stage means for it, the areas it serves, and how they depend on each other.
Commercial Products
Enterprise software pipeline
| Code | Name | Stage | Progress | Areas | Note | Description | Links |
|---|---|---|---|---|---|---|---|
| SCI2-COM-VC-0001 | VectorCat | Phase III (Proven) (stage 6 of 7) | Completed: Discovery (Prototype), Lead Gen (MVP), Safety (Alpha/Beta), Phase I (Release), Phase II (Scaling). Active: Phase III (Proven). Upcoming: Industry Standard. | Data Governance, Knowledge Engineering, Research Assistance, AI Safety | Multiple sites, efficacy measurements from open science repos and internal deployments. Currently measuring enterprise-wide deployment sites. | Enterprise knowledge platform — catalog, search, analysis, AI inference, agent fleet across regulated industries. Currently measuring enterprise-wide deployment sites. | Product page |
| SCI2-COM-IC-0001 | Integrity Cortex | Phase I (Release) (stage 4 of 7) | Completed: Discovery (Prototype), Lead Gen (MVP), Safety (Alpha/Beta). Active: Phase I (Release). Upcoming: Phase II (Scaling), Phase III (Proven), Industry Standard. | Data Integrity, AI Safety, Formal Methods, Research Assistance | Organising first sites with partners | Neurosymbolic memory layer — every claim proved, not promised. Cascade dependency tracking across the enterprise | Product page |
| SCI2-COM-SN-0001 | Salience Network | Lead Gen (MVP) (stage 2 of 7) | Completed: Discovery (Prototype). Active: Lead Gen (MVP). Upcoming: Safety (Alpha/Beta), Phase I (Release), Phase II (Scaling), Phase III (Proven), Industry Standard. | Portfolio & Project Mgmt, Stochastic Process Modeling, Creative Problem Solving, Formal Methods, Knowledge Engineering, AI Safety | Active development — pipeline visualization, cross-dependency mapping | Portfolio orchestration — staged pipeline tracking, cross-dependency mapping, roadmap governance. Brainwave-phase synchronization for the enterprise | In development |
| SCI2-COM-CB-0001 | ▒▒▒▒▒▒▒▒ | In stealth | In stealth | ||||
| SCI2-COM-PMC-0001 | PMChat | Phase I (Release) (stage 4 of 7) | Completed: Discovery (Prototype), Lead Gen (MVP), Safety (Alpha/Beta). Active: Phase I (Release). Upcoming: Phase II (Scaling), Phase III (Proven), Industry Standard. | Research Assistance, Knowledge Engineering | First research teams onboarded | AI research assistant — multi-provider chat with PubMed/bioRxiv integration and paper management for life sciences | pmchat.app |
| SCI2-COM-SRP-0001 | Serpents | Phase I (Release) (stage 4 of 7) | Completed: Discovery (Prototype), Lead Gen (MVP), Safety (Alpha/Beta). Active: Phase I (Release). Upcoming: Phase II (Scaling), Phase III (Proven), Industry Standard. | Software & Dev Tools, Workflow Orchestration, AI Safety, Knowledge Engineering | Deployed in VectorCat | DAG pipeline engine — LangChain of VectorCat, Integrity Cortex and Parseltongue, implementing SAFE framework | AI Agents |
| SCI2-COM-VBP-0001 | Vibes-PRO | Lead Gen (MVP) (stage 2 of 7) | Completed: Discovery (Prototype). Active: Lead Gen (MVP). Upcoming: Safety (Alpha/Beta), Phase I (Release), Phase II (Scaling), Phase III (Proven), Industry Standard. | Software & Dev Tools | Pipeline components, advanced data viz | Proprietary design system layer — pipeline visualization, advanced data components, domain-specific widgets built on Vibes-OSS | In development |
Open Source
Tools and libraries
| Code | Name | Stage | Progress | Areas | Note | Description | Links |
|---|---|---|---|---|---|---|---|
| SCI2-OSS-PT-0001 | Parseltongue DSL | Phase III (Proven) (stage 6 of 7) | Completed: Discovery (Prototype), Lead Gen (MVP), Safety (Alpha/Beta), Phase I (Release), Phase II (Scaling). Active: Phase III (Proven). Upcoming: Industry Standard. | Formal Methods, Data Integrity, AI Safety | OSSed, external users, deployed across VectorCat, Integrity Cortex, PMChat, Parseltongue LLM, Bench, TUI | Formal-reasoning language — evidence-grounded facts, axioms, derivations, diffs, taint propagation. The language that refuses to speak falsehood | PyPI, GitHub |
| SCI2-OSS-PT-0002 | Parseltongue Bench | Phase I (Release) (stage 4 of 7) | Completed: Discovery (Prototype), Lead Gen (MVP), Safety (Alpha/Beta). Active: Phase I (Release). Upcoming: Phase II (Scaling), Phase III (Proven), Industry Standard. | Software & Dev Tools, Open Science, Data Integrity, Knowledge Engineering | Persistent daemon with full-text search, D3 viz, notebooks | Agentic inspection server — persistent daemon, full-text s-expression search, D3 visualization, interactive notebooks | GitHub |
| SCI2-OSS-PT-0003 | Parseltongue LLM | Safety (Alpha/Beta) (stage 3 of 7) | Completed: Discovery (Prototype), Lead Gen (MVP). Active: Safety (Alpha/Beta). Upcoming: Phase I (Release), Phase II (Scaling), Phase III (Proven), Industry Standard. | AI Safety, Research Assistance | Internal use — synthetic model testing | 4-pass LLM pipeline — extract → blinded derivation → fact-check → grounded answer. Document analysis via Docling (PDF/DOCX/PPTX) | GitHub |
| SCI2-OSS-PT-0004 | Parseltongue TUI | Safety (Alpha/Beta) (stage 3 of 7) | Completed: Discovery (Prototype), Lead Gen (MVP). Active: Safety (Alpha/Beta). Upcoming: Phase I (Release), Phase II (Scaling), Phase III (Proven), Industry Standard. | Research Assistance, Open Science | Internal use — daily driver | Interactive terminal app — run multi-document analysis, inspect formal systems, browse history. The human interface to Parseltongue | GitHub |
| SCI2-OSS-XEN-0001 | xen | Safety (Alpha/Beta) (stage 3 of 7) | Completed: Discovery (Prototype), Lead Gen (MVP). Active: Safety (Alpha/Beta). Upcoming: Phase I (Release), Phase II (Scaling), Phase III (Proven), Industry Standard. | Software & Dev Tools | Internal use, manages our virtual monorepo including this project | Virtual monorepo — open portals, summon repos, cascade commands across the stack. Intent-based workspace without submodules | GitHub |
| SCI2-OSS-VB-0001 | Vibes-OSS | Lead Gen (MVP) (stage 2 of 7) | Completed: Discovery (Prototype). Active: Lead Gen (MVP). Upcoming: Safety (Alpha/Beta), Phase I (Release), Phase II (Scaling), Phase III (Proven), Industry Standard. | Software & Dev Tools | React/TypeScript component library, Catppuccin theming | Open-source design system — React/TypeScript components, Catppuccin theming, Tailwind tokens. The base layer Vibes-PRO extends | In development |
Research
Papers and publications
| Code | Name | Stage | Progress | Areas | Note | Description | Links |
|---|---|---|---|---|---|---|---|
| SCI2-RES-SF-0001 | SAFE Framework | Preprint (stage 3 of 5) | Completed: Hypothesis, Draft. Active: Preprint. Upcoming: Under Review, Published. | Data Integrity, Data Governance, AI Safety, Open Science | Published on sci2sci.com | Secure & Auditable FAIR Environment — extends FAIR principles with zero-trust security and grounded audit evidence for regulated biopharma | SAFE Principles |
| SCI2-RES-VS-0001 | V* (Critical Volatility) | Under Review (stage 4 of 5) | Completed: Hypothesis, Draft, Preprint. Active: Under Review. Upcoming: Published. | Stochastic Process Modeling, Open Science | Under peer review | Paper describing stochastic process criticality — derives phase-transition thresholds for log-normal iterative processes | GitHub |
| SCI2-RES-TBP-0001 | Two-Board Problem | Preprint (stage 3 of 5) | Completed: Hypothesis, Draft. Active: Preprint. Upcoming: Under Review, Published. | Creative Problem Solving, Open Science | arXiv preprint — MDP training environment | Paper formalizing a mathematical learning challenge — agent must discover algebraic structures with zero intermediate reward signal | GitHub |
| SCI2-RES-OA-0001 | Oracles & Axioms | Preprint (stage 3 of 5) | Completed: Hypothesis, Draft. Active: Preprint. Upcoming: Under Review, Published. | Formal Methods, Open Science | Draft finalized, preprint pending | Paper proving equivalence between inaccessible rewrite rules in computer algebra and Turing oracles — any rule external to decidable axioms functions as a true oracle | GitHub |
| SCI2-RES-QE-0001 | Quantum Epistemics | Draft (stage 2 of 5) | Completed: Hypothesis. Active: Draft. Upcoming: Preprint, Under Review, Published. | Formal Methods, Open Science | Past hypothesis, draft in progress | Epistemics formalization — superposed states and lattice theory for modelling observer states | In development |
Dependencies
| From | To | Relation |
|---|---|---|
| Parseltongue DSL · Lead Gen (MVP) | Parseltongue Bench · Lead Gen (MVP) | DSL is the core of Parseltongue Bench |
| Parseltongue DSL · Lead Gen (MVP) | Parseltongue LLM · Lead Gen (MVP) | DSL powers Parseltongue LLM extraction pipeline |
| Parseltongue DSL · Lead Gen (MVP) | Parseltongue TUI · Lead Gen (MVP) | DSL powers Parseltongue TUI |
| Parseltongue DSL · Phase I (Release) | Integrity Cortex · Lead Gen (MVP) | Parseltongue DSL powers Integrity Cortex formal verification |
| Parseltongue DSL · Phase I (Release) | PMChat · Phase I (Release) | PMChat depends on Parseltongue DSL |
| Parseltongue DSL · Phase II (Scaling) | VectorCat · Phase II (Scaling) | VectorCat uses Parseltongue for knowledge verification |
| Parseltongue LLM · Lead Gen (MVP) | Parseltongue TUI · Lead Gen (MVP) | TUI wraps the LLM pipeline |
| Vibes-OSS · Lead Gen (MVP) | Vibes-PRO · Lead Gen (MVP) | Vibes-PRO extends Vibes-OSS |
| Vibes-OSS · Lead Gen (MVP) | Parseltongue Bench · Phase II (Scaling) | Vibes-OSS powers Parseltongue Bench Phase II frontend |
| Vibes-PRO · Lead Gen (MVP) | VectorCat · Phase II (Scaling) | Vibes-PRO powers VectorCat Phase II frontend |
| Vibes-PRO · Lead Gen (MVP) | PMChat · Phase I (Release) | Vibes-PRO powers PMChat Phase I frontend |
| Vibes-PRO · Lead Gen (MVP) | Salience Network · Lead Gen (MVP) | Vibes-PRO pipeline components power Salience Network development |
| Serpents · Phase I (Release) | VectorCat · Phase II (Scaling) | Serpents DAG engine powers VectorCat pipelines |
| Serpents · Phase II (Scaling) | PMChat · Phase II (Scaling) | Serpents planned for PMChat Phase II |
| Parseltongue DSL · Phase III (Proven) | Serpents · Phase II (Scaling) | Serpents Phase II natively speaks Parseltongue |
| VectorCat · Phase I (Release) | Integrity Cortex · Lead Gen (MVP) | VectorCat Phase I is the base for Integrity Cortex development |
| Integrity Cortex · Safety (Alpha/Beta) | ▒▒▒▒▒▒▒▒ · In stealth | Integrity Cortex feeds coherence state to downstream |
| Integrity Cortex · Safety (Alpha/Beta) | Salience Network · Lead Gen (MVP) | Integrity Cortex safety feeds Salience Network development |
| V* (Critical Volatility) · Preprint | Salience Network · Discovery (Prototype) | V* stochastic modeling research feeds Salience Network |
| Two-Board Problem · Preprint | Salience Network · Discovery (Prototype) | Two-Board Problem MDP research informs Salience Network design |
| Oracles & Axioms · Draft | Parseltongue DSL · Lead Gen (MVP) | Oracles & Axioms originates Parseltongue as a project |
| SAFE Framework · Preprint | VectorCat · Phase I (Release) | SAFE framework defines VectorCat Phase I security and auditability model |
| SAFE Framework · Preprint | Serpents · Lead Gen (MVP) | SAFE principles implemented natively in Serpents pipeline engine |
| Parseltongue DSL · Phase I (Release) | Quantum Epistemics · Draft | Quantum Epistemics uses Parseltongue Std Lib epistemics definitions |
| Parseltongue DSL · Safety (Alpha/Beta) | Parseltongue Bench · Safety (Alpha/Beta) | DSL safety validates Bench safety |
| Parseltongue DSL · Safety (Alpha/Beta) | Parseltongue TUI · Safety (Alpha/Beta) | DSL safety validates TUI safety |
| Parseltongue DSL · Safety (Alpha/Beta) | Parseltongue LLM · Safety (Alpha/Beta) | DSL safety validates LLM pipeline safety |
| Parseltongue DSL · Phase I (Release) | Parseltongue Bench · Phase I (Release) | DSL Phase I enables Bench Phase I deployment |
| Parseltongue DSL · Phase III (Proven) | V* (Critical Volatility) · Published | Published as Knowledge As Code via Parseltongue |
| Parseltongue DSL · Phase III (Proven) | Two-Board Problem · Published | Published as Knowledge As Code via Parseltongue |
| Parseltongue DSL · Phase III (Proven) | Oracles & Axioms · Published | Published as Knowledge As Code via Parseltongue |
| Parseltongue DSL · Phase III (Proven) | Quantum Epistemics · Published | Published as Knowledge As Code via Parseltongue |
Why stages matter
While our research tracking may look familiar, you may have noticed we approach software maturity with more rigor than a typical SaaS. If we advertise that a product helps your use cases, we provide sufficient proof and are transparent about how strong evidence currently is.
To give you some intuition of why we decided to extend the typical framework: "released" software highly varies in its maturity. A working product with no controlled evidence that it solves the problem at scale — we call this Lead Gen (MVP). Think early Notion, Linear or most AI startups today. A vendor with first real customers and "social proof" testimonials but no controlled measurements — that's Phase I (Release). Software that has become the go-to way to resolve its use case — Medidata for clinical data capture, ANSYS for engineering simulation, SAS for statistical computing — is what we call Industry Standard. All three are "released" in the conventional sense, but the distance between them is enormous.
Our philosophy: we price according to the stage. Early adopters who join at Phase I or Phase II lock in multi-year discounts that persist regardless of how fast we progress. For OSS products, the same applies to the controlled deployment and support tiers.
Product stages
Each gate below is the minimum bar to enter the next stage, not a marketing milestone. A landing page with no research behind it doesn't even reach Discovery. An enterprise product with 100M data objects processed and measured time savings is still not an Industry Standard.
| Stage | Meaning | Gate to next |
|---|---|---|
| Discovery (Prototype) | Research-backed problem framing, feasibility prototype | Validated hypothesis, decision to build |
| Lead Gen (MVP) | Core architecture, first working product | Functional MVP deployable internally |
| Safety (Alpha/Beta) | Controlled deployment (internal or external alpha/beta), synthetic and real-world stress testing | No critical failures under controlled conditions |
| Phase I (Release) | First external deployment, real users, real data | Stable in production, no safety incidents |
| Phase II (Scaling) | Multiple deployment sites, first controlled efficacy measurements | Quantified outcomes across independent sites |
| Phase III (Proven) | Large-scale deployments, statistical significance of customer KPI improvements | Meets criteria for industry standard |
| Industry Standard | The product is an industry-standard way for resolving the use case | — |
Research stages
| Stage | Meaning |
|---|---|
| Hypothesis | Idea, early exploration |
| Draft | Writing in progress |
| Preprint | Published on arXiv / bioRxiv, not peer-reviewed |
| Under Review | Peer review in progress |
| Published | Peer-reviewed, published in a trusted source |