3 min read
On this page
Roadmap: ShunyaBar Labs, H2 2026
Based on CPO Product Strategy v2 (June 2026). Primary: Wedge A (Verifiable MaxSAT API). Secondary: Wedge B (ZK Pre-Processor). Kill: Wedge C.
Vault Door Principle: ZK should stop being the front door. It should become the vault door. ZK stays in the long-term moat. We demote lead-with-ZK framing on the homepage, not ZK itself. Cutting ZK from the company thesis would turn a differentiated proof-oriented solver into a generic solver API competing against Z3 and MaxHS.
Executive Snapshot#
For a reader with 60 seconds: This is what we ship, in what order, against what targets.
| Quarter | What we ship | Revenue target | First-time hires |
|---|---|---|---|
| Q3 2026 | MaxSAT-first homepage; fixed API schema (integer weights, derivation graphs, infeasibility case); Dev $99 / Pro $499 pricing locked | $20K | Solutions Engineer (week 1) |
| Q4 2026 | ZK pre-processor sidecar; proof-of-optimality artifacts in API; open-source verifier (Apache 2.0) | $80K cumulative | DevRel, Backend Engineer |
| Q1 2027 | Enterprise tier (on-prem, SOC2 in flight); open-source engine core; first paid ZK audit pilot (~$25K) | $250K cumulative | ZK Engineer (only if wedge validated) |
| Q2 2027 | ZK product-line decision gate — 100% verifiable solver + <1s p99 + one paying ZK customer required | $500K cumulative | — |
The single thing that matters this quarter: 3 signed design partner LOIs by Sept 30. Without these, the rest is feature roadmap.
What we are NOT doing: ZK DSL, generic SAT, mobile/edge, custom hardware, Wedge C (agent guardrails).
Critical Fixes (Schema Bugs)#
- Integer weights only — float weights invalidate dual-bound theory. Document this in API docs.
- Derivation graphs, not strings —
resolution_chainmust be structured (clause IDs + resolution steps), not English text. - Infeasibility case — Add
satisfiable: falseresponse with hard-clause UNSAT core and derivation. Clients must distinguish “optimal” from “infeasible” from “engine gave up.”
Phase 1: Q3 2026 — Revenue First#
Human Moves (Priority Over Engineering)#
- 3 design partner LOIs signed by Sept 30
- 1 scheduling / timetabling ($5K–$20K)
- 1 ML feature selection ($5K)
- 1 ZK preprocessor evaluation (free → paid)
- Pilot template ready — problem, success criteria, timeline, payment terms, IP clause
- First hire: Solutions Engineer — starts Q3 week 1
- 30-day mandate: prospect list, pilot template, 10 outreach calls
Marketing & Positioning#
- Demote ZK on homepage (vault door, not front door) — MaxSAT-first positioning. ZK moves to proof layer and product-strategy copy, not the headline. Don’t amputate the moat.
- Rewrite homepage copy: “Verifiable MaxSAT API” not “constraint intelligence”
- Create Wedge A landing page (Verifiable MaxSAT)
- Wedge B teaser page (“ZK Pre-Processor — coming Q4”)
Product#
- Fix API schema — integer weights, derivation graphs, infeasibility case
- Dev tier: $99/mo (10K vars, 50K clauses, 5K req/hr)
- Pro tier: $499/mo (100K vars, 500K clauses, 50K req/hr)
- Free tier: 100 req/hr → 500 req/hr
Engineering#
- Audit mini engine correctness — 99.77% → 99.9%+? If not, lower the number honestly.
- Stabilize zk_binary_v2.py SDK
Research#
- Submit EA 3D paper (Physical Review E or ML venue)
ARR target: $20K by Sept 2026
Phase 2: Q4 2026 — Ship Wedge B + Proof Artifacts#
Product#
- ZK Pre-Processor sidecar — free decomposition, paid re-encoding optimizer
- Per-circuit pricing: $50–$500
- Partner integration: snarkjs or gnark “use NitroSAT as preprocessor”
- Proof-of-optimality artifact in API response (UNSAT cores + derivation graph)
Open Source#
- Open-source verifier + proof format spec (Apache 2.0)
- Proof format JSON schema
- Verifier implementation
- Example clients
- Download + run offline = independent verification
Engineering#
- Optimize /v1/schedule for monthly/quarterly horizons
- On-prem Docker container (beta)
Hiring#
- Second hire: DevRel — open-source community, SDK docs, 3 posts
ARR target: $80K cumulative by Dec 2026
Phase 3: Q1 2027 — Enterprise + Open Core#
Product#
- Enterprise tier launch — on-prem + SOC2 in flight
- Open-source engine core (Apache 2.0) — parsing, CNF/XOR, basic local-search loop
- Closed layer: proof generation tuning, telemetry, RBAC, audit logging, ZK heuristics
- Proof artifact verification library (open source)
Engineering#
- Metered billing API
- SSO & RBAC
- Audit logs
Sales#
- First paid ZK audit firm pilot (~$25K contract)
ARR target: $250K cumulative by Mar 2027
Phase 4: Q2 2027 — ZK Product-Line Decision Gate#
ZK is in the stack from day one as the long-term trust and settlement layer. This gate validates whether ZK earns a separate product line, not whether ZK is in the company.
ZK Spin-Out Decision (Requires ALL of):#
- 100% verifiable solver (mini audit: 40-60% there today)
- <1s p99 latency on depth-4-equivalent
- At least one paying ZK customer in production
If yes: Spin out “Shunyabar ZK” as separate product line If no: Stay MaxSAT-only; revisit spin-out in 2028. ZK remains in the verification layer regardless.
Hiring (Full Team)#
- Solutions Engineer (Q3 — first hire)
- DevRel (Q4)
- Backend Engineer (Q4)
- ZK Engineer (Q1 2027 — only if ZK wedge validated)
ARR target: $500K cumulative by Jun 2027
Kill List (12 months)#
| Do NOT do | Why |
|---|---|
| Real-time crypto / wallet / signing | Latency + audit surface |
| Build ZK DSL | Circom/gnark/Noir own this |
| General SAT market | Z3/Kissat/CaDiCaL dominate |
| Mobile / edge deployment | Engine footprint too big |
| Custom hardware (GPUs) | Research play, not 2027 product |
| Lead-with-ZK framing on homepage | Until paying ZK customer exists |
| Wedge C (Agent Guardrails) | CENTS test failed — kill it |
| Hire ZK engineer in Q3 | Premature — ZK wedge not validated |
Do NOT kill (Preserve)#
| ✅ Preserve | Why |
|---|---|
| ZK in the company thesis and verification roadmap | Long-term moat. Cutting it turns a proof-oriented solver into a generic solver API. |
| ZK as the trust layer in the proof artifact | What makes “verifiable MaxSAT” defensible. |
| Wedge B (ZK Pre-Processor) | Validated as the 2027+ enterprise audit sidecar. |
| ZK product-line decision gate in Q2 2027 | Validates whether ZK earns a product line, not whether ZK is in the company. |
Current vs Target Pricing#
| Tier | Current | Target |
|---|---|---|
| Free | $0 (100 req/hr) | $0 (500 req/hr) |
| Dev | — | $99/mo |
| Pro | $0.25 + $0.10/min | $499/mo |
| Enterprise | $100/mo + $10/hr | Custom $5K+/mo |
What We Shipped (Pre-Strategy)#
| Item | Status |
|---|---|
| MATH.md (886 lines, rigorous) | Done |
| Limitations & Boundaries page | Done |
| Remove “beats Kissat” claims | Done |
| Fix 2^1024 XOR framing | Done |
| CDCL honest comparison (anytime vs exact) | Done |
| Industrial-Scale results (80M-clause, EA 3D, MSE 2022) | Done |
| CSS improvements (code, tables, math, links, sidebar) | Done |
| Project pages with live URLs | Done |
| Product Strategy document | Done |
| CENTS test table | Done |
| Proof-of-optimality API schema (integer weights, derivation graphs, infeasibility case) | Done |
| Corrected open-source split (open verifier, closed proof generation) | Done |
| Hiring sequence + pilot template | Done |
Revenue first. LOIs before artifacts. Open-source verifier. Ship the vehicle, not just the engine.
See Also#
- Product Strategy — the strategic document behind this roadmap
- The Bottleneck — funnel audit with 90-day targets
- Pricing — current vs target pricing
- Marketing: Why Navokoj · Use Cases · Results
- Getting Started — what’s shippable today