All workPrivate beta

Founder case study · discovery, product model, delivery and evaluation

Orchestra

Orchestra separates immutable project evidence from reviewed product truth, turning PRDs, meetings, client conversations and code into citation-backed answers, proposed requirement changes and approved context for teams and AI coding tools.

Discovery
100+ customer interviews
Validation
Five pilot customers
Evaluation
148 retrieval and grounding cases
Orchestra Socrates workspace for grounded questions, citations and evidence controls
Socrates product workspace using deterministic demonstration data.Synthetic demonstration data

The recurring problem was not search. It was lost product context.

Across more than 100 customer interviews, the same failure pattern kept appearing. Decisions lived in meetings, client conversations, PRDs, tickets and code. Teams could find individual artefacts, but they could not reliably reconstruct why the product had changed or which statement should guide the next decision.

That changed the product thesis. Orchestra would not be another chat layer over documents. It would preserve evidence, surface conflicts and create an explicit review path before a proposal could become shared product truth.

From searchable memory to a governed product brain

  1. DiscoveryMap fragmented decisions and repeated context loss.
  2. RetrievalUnify approved sources with provenance-aware search and reranking.
  3. Truth modelSeparate immutable evidence from reviewed, versioned product truth.
  4. DeliveryExpose grounded answers in Socrates and approved context through 17 read-only integrations.
  5. ValidationRun the workflow with five pilot customers and protect behaviour with 148 cases.

Evidence is not truth.

A meeting note, client message or pull request can be authentic evidence and still conflict with the current product direction. Orchestra therefore never silently promotes retrieved text into accepted context.

01Preserve evidenceImmutable source, provenance and permissions
02Propose a changeQuestion, conflict or requirement delta
03Review the mappingHuman accepts, rejects or edits the proposal
04Version product truthApproved context becomes auditable state

One control path from source to answer

Retrieval finds relevant evidence. Typed workflow logic and review controls decide what may change. Socrates and MCP deliver only the context appropriate to the requesting user and surface.

  1. Sources

    PRDs, meetings, conversations, GitHub and Drive

  2. Immutable evidence

    Source records preserved with provenance and permissions

  3. Hybrid retrieval

    Dense and sparse search with metadata filters and reranking

  4. Review proposal

    Questions, conflicts and requirement changes mapped for review

  5. Versioned truth

    Accepted changes update auditable Product Brain state

  6. Socrates and MCP

    Grounded answers and approved context delivered where work happens

The governance model is visible in the interface.

Ask

Socrates

Grounded answers expose citations and make low-evidence states explicit.

Review

Live Doc

Proposed requirement changes remain inspectable before accepted context moves.

Deliver

MCP and integrations

Approved context reaches tools through read-only boundaries rather than hidden write access.

The suite tests behaviour, not presentation.

The public portfolio claim remains the résumé-verified 148-case suite. The categories below describe what that evaluation strategy protects.

What the pilots validated, and what remains.

Validated

Teams value a shared path from scattered evidence to grounded answers and reviewable product decisions.

Still constrained

The beta remains private, customer data is excluded from this portfolio and public visuals use deterministic demonstration records.

My ownership

Customer discovery, Product strategy, System architecture, Retrieval and evaluation design, Engineering delivery, Pilot execution.