Book demo

Founder-led roadmap

A roadmap by release, not by artificial dates.

Nolta moves in a Tikk / Takk rhythm. A Tikk is a major functionality release: new product capability, new context loop, new surface or new integration layer. A Takk is the stability release that follows: hardening, entitlement checks, UI consistency, reliability, tests and cleanup.

Each release is backed by a Markdown file. Released versions describe what actually shipped. Future versions describe the intended headlines without pretending that early product discovery can be scheduled down to the day.

Book a demo →Context

Release sequence

What each version is meant to do.

The roadmap starts with the future direction at the top, then moves backward through planned, in-development and released versions. No dates are attached here on purpose.

Direction · Exploration

Later — Platform expansion after the core is calm

Important directions that fit Nolta, but should not distract from making the core product excellent first.

Direction

  • Focused connectors where customers need them, built around Nolta's context-first integration philosophy.
  • Controlled adapters for AI tools and customer systems where they improve real workflows.
  • Advanced analytics, retrospectives and heat-over-time views.
  • Enterprise-readiness evidence and stronger operating habits.
  • Optional focused AI workspaces around a node, timeline moment or project scope.
  • Carefully sanitized demo worlds for selected vertical conversations.

Product meaning

Nolta should grow as living context infrastructure: useful for humans first, powerful for AI because the context is already structured, scoped and reviewed.

Tikk · Exploration

v1.4 — Platform efficiency and operational evidence

A platform-oriented direction for reducing AI waste, improving context quality and connecting reviewed work to relevant time-stamped operational evidence.

Release theme

v1.4 explores how Nolta can help teams use AI more efficiently while keeping work explainable.

The public product idea:

**Send better context. Waste fewer tokens. Preserve the trace.**

Direction

### Context efficiency

### Reviewable context reduction

### External operational evidence

### Platform and adapter maturity

### Trust and execution analytics

  • Help teams reduce irrelevant context before AI-assisted work runs.
  • Preserve traceability and human review while making context smaller and more useful.
  • Prefer explainable scoping, summaries and references over large prompt dumps.
  • Keep detailed compaction mechanics private until they are protected, validated and customer-ready.
  • Show users what was kept, reduced, summarized or referenced.
  • Preserve the ability to inspect source material when needed.
  • Use product analytics to identify context that is often irrelevant, stale or repeatedly removed during review.
  • Explore a general way to attach time-stamped external observations to Nolta context.
  • Treat external observations as supporting context, not as automatic project truth.
  • Preserve source, time, freshness and confidence where external signals are used as evidence.
  • Keep public examples broad and synthetic until partner-specific use cases are agreed.
  • Continue improving Nolta as a controlled context layer around existing tools and AI systems.
  • Keep integrations scoped, permission-aware and review-oriented.
  • Avoid presenting every possible adapter or protocol as a committed public promise before customer demand is clear.
  • Expand organization-level insight around context quality, review flow, AI cost and outcomes.
  • Recommend improvements when review queues, weak evidence or unnecessary context create friction.
  • Keep analytics focused on process health and trust quality, not individual surveillance.

Product meaning

v1.4 should make Nolta feel like a serious platform for AI-era context operations.

The product should help companies send better context, preserve evidence and explain why work happened the way it did.

This is where Nolta's value becomes sharper: not just “we have context,” but “we can prepare the right context, review proposed work and remember the approved change.”

Boundary

In scope:

Out of scope unless explicitly reprioritized:

  • Context efficiency and reviewable reduction.
  • Safer use of supporting operational evidence.
  • Continued platform and integration maturity.
  • Organization-level analytics for trust, cost and context quality.
  • Publicly claiming proprietary AI execution mechanics before protection decisions are made.
  • Broad autonomous-agent positioning.
  • Replacing human review with automatic truth acceptance.
  • External observations silently overwriting trusted project context.
  • Broad enterprise BI unrelated to context, review, cost or evidence quality.
Tikk · Exploration

v1.3 — Search, evidence and attention

A release direction for helping teams navigate more context with less noise: better search, stronger evidence surfaces, calmer review queues and useful attention signals.

Release theme

v1.3 should make Nolta easier to use when real context starts to grow.

The core idea is simple: Nolta should help users find the surrounding context behind work, not just matching records or isolated files.

Direction

### Context search

### Attachment-aware evidence

### Better review and attention surfaces

### Timeline intelligence and filters

### Push-based attention

### Analytics foundations

  • Improve search so users can find the relevant project, node, moment, decision, risk or review trail behind a question.
  • Combine semantic retrieval with Nolta's existing structure, permissions, trust state and timeline context.
  • Keep canonical truth in Nolta's main records; retrieval remains an aid, not a replacement for the source of truth.
  • Preserve the public framing: search should return context, not just snippets.
  • Make uploaded material easier to search, reference and review where customer permission allows it.
  • Return matches with surrounding Nolta context: where the material was added, what it relates to and what else was happening at the time.
  • Let reviewed work point back to supporting material without overclaiming what an AI model internally relied on.
  • Keep extraction limitations visible and honest.
  • Create calmer surfaces for pending proposals, unresolved risks, findings, verification needs and recent changes.
  • Help reviewers understand what deserves attention without turning Nolta into a noisy task feed.
  • Show proposed changes, supporting evidence and expected impact before approval.
  • Make it easier to understand what changed, when it changed and why it mattered.
  • Use filters and visual emphasis to reshape context around role, state, time, heat and attention.
  • Support dated review moments, reminders and important context markers.
  • Reduce unnecessary polling by moving toward event-driven UI updates where appropriate.
  • Keep notification state reliable across reconnects and missed events.
  • Use attention signals for review requests, completed work, failed verification and newly important findings.
  • Add early organization-level insight around context quality, review bottlenecks and AI-assisted work outcomes.
  • Focus on process improvement, not employee surveillance.
  • Use recommendations to help teams reduce irrelevant context, improve review capacity and keep evidence strong.

Product meaning

v1.3 should make Nolta feel sharper and calmer as the amount of context grows.

Users should be able to ask: “What matters here?”, “Where did this come from?”, “Which evidence supports this?”, “Who needs to review this?” and “What changed since last time?”

The answer should be grounded in Nolta context, not hidden inside search snippets or AI chat logs.

Boundary

In scope:

Out of scope unless explicitly reprioritized:

  • Better context search.
  • Attachment-aware search and evidence surfaces.
  • Stronger dashboard, review and attention views.
  • Timeline filters and context markers.
  • Event-driven attention updates where useful.
  • Early trust and execution analytics.
  • Replacing the canonical data model with a search index.
  • General web crawling.
  • Broad enterprise BI.
  • Personal employee surveillance.
  • Heavy realtime/event-streaming infrastructure before it is needed.
  • Fully autonomous agent orchestration.
Tikk · In Development

v1.2 — Trust & Execution

The release that makes Nolta's AI-assisted work loop clearer, safer and easier to review: scoped context, governed proposals, human approval and remembered trust.

Release theme

v1.2 is the trust-spine release for Nolta.

The goal is not autonomous magic. The goal is controlled AI-assisted work: Nolta helps teams prepare relevant context, review proposed changes, approve what is useful and preserve the trace afterward.

The public product sentence for this release is simple:

**AI proposes. Humans approve. Nolta remembers.**

Planned direction

### Node-centered Trust & Execution flow

### Governed execution boundaries

### Assisted universe creation

### Nicky as context intelligence

### Reviewable context preparation

### Structured proposals and approved action

### Provider and cost clarity

### Observatory trust markers

### Product confidence

  • Make the normal AI-assisted work loop easier to run from the place where the context already lives.
  • Keep review, approval, verification and trust state visible without forcing users through disconnected screens.
  • Preserve a clear return path from any proposal or execution detail back to the original context.
  • Add clearer checks before AI-assisted work begins.
  • Explain whether the selected context is ready, whether the requested work fits the current scope and what kind of review will be needed.
  • Prevent broad, irrelevant or poorly supported AI work from running from the wrong place.
  • Keep the public message focused on safety, scope and review rather than implementation mechanics.
  • Let Nolta help turn source material such as documents, plans, research notes, creative briefs or technical scope into an initial reviewed context universe.
  • Keep humans in charge of structure choices when the source can reasonably be interpreted in more than one way.
  • Record the difference between what the assistant proposed and what the user approved.
  • Use Nicky to inspect context quality, structure, gaps, unclear ownership and weak evidence.
  • Keep Nicky's role advisory: it can find, explain and propose, but trusted context still requires human approval.
  • Present findings in a way that helps teams improve context rather than drown in alerts.
  • Make it easier to understand what context is being used for a task and why.
  • Help users remove irrelevant material and keep the work scoped.
  • Preserve source, scope and evidence information for later review without exposing private implementation details publicly.
  • Treat AI output as a proposal, not as automatic truth.
  • Support item-level review where needed, including accept, edit, reject or request-more-work decisions.
  • Keep approved changes traceable to the context and review that produced them.
  • Avoid pretending human review is binary; Nolta should remember how strongly something was reviewed.
  • Make provider choice, privacy boundaries and cost expectations understandable to users and admins.
  • Avoid silent changes in how AI work is routed.
  • Keep deeper orchestration details private until customer and partner needs justify exposing them.
  • Make trust, review and verification state visible in the Galaxy/Observatory experience.
  • Reuse the same visual language across detail views and the broader context map.
  • Keep attention markers useful without mixing human notes and system trust state into one confusing layer.
  • Harden the core trust loop before expanding into broader automation.
  • Strengthen testing, regression protection and demo-readiness around context creation, review, approval and traceability.
  • Keep the public demo experience polished, read-only where appropriate and safe to share.

Product meaning

v1.2 should make Nolta feel like the approval and context layer between AI reasoning and real-world change.

Nicky should feel like governed context intelligence: part inspector, part guide, part reviewer of context quality. It should not feel like a chatbot bolted onto project data.

AI can inspect, explain and propose. Humans approve. Nolta remembers the trace.

Boundary

v1.2 is about building the spine, not every future capability.

In scope:

Out of scope unless explicitly reprioritized:

  • A clearer Trust & Execution loop.
  • Safer context preparation and proposal review.
  • Human-approved changes with traceability.
  • Assisted universe creation with review.
  • Context-quality findings and advisory recommendations.
  • Better visual trust markers.
  • Product hardening and demo confidence.
  • Fully autonomous long-running agents.
  • Broad autonomous orchestration.
  • Direct external system writeback as the public default.
  • General web crawling.
  • Deep vertical-specific operational integrations.
  • Compliance claims beyond what the product and process can actually support.
Takk · Released

v1.1.1 — Product hardening after v1.1

The stabilization release after the first end-to-end context loop: clearer UI, cleaner packaging, stronger confidence and better product readiness.

Release theme

v1.1.1 was a calm-down release after the fast v1.1 push.

The goal was to make the core easier to understand, easier to test and safer to demonstrate before Nolta moved deeper into Trust & Execution.

Shipped direction

  • Cleaned up the core user experience around context, review and proposal details.
  • Improved visual consistency across Galaxy, Rethread and detail surfaces.
  • Strengthened build, smoke-test and release confidence.
  • Polished licensing and deployment readiness without exposing internal mechanics publicly.
  • Improved demo stability and reduced product rough edges.

Product meaning

The product message is simple: the core stays useful and deterministic, while AI-assisted context features remain controlled and reviewable.

v1.1.1 is the bridge into v1.2. Before Nolta promises deeper Trust & Execution, the current foundation needs to be solid.

Tikk · Released

v1.1 — The first context-to-review loop

The release where Nolta became more than a project view: selected context, AI-assisted proposals, human review and timeline memory started to connect into one loop.

Release theme

v1.1 introduced the first practical loop between Nolta context, AI-assisted work, review and remembered change.

Shipped direction

  • A foundation for preparing selected context for AI-assisted work.
  • Reviewable proposals instead of direct AI mutation.
  • Context-related activity visible in the project thread and timeline.
  • Nicky findings built around inspectable context quality signals.
  • Early licensing and product-boundary foundations.

Product meaning

v1.1 proved the core idea: AI work should begin from context, return as a proposal, pass through human review and remain traceable afterward.