| Posture |
| Primary use case | Buyer-facing CPQ slice: discovery through validated configuration, pricing, and proposal — governed quote record may live elsewhere. Catalog and commercial rules can be modeled in Talkulate AI CPQ AI CPQ's database without a Smart CPQ tenant when Talkulate AI CPQ-only is the program. | PROS / Conga Smart CPQ is enterprise rep- and deal-desk-centric CPQ with constraints-based configuration, CRM one-way sync, and suite adjacency to price optimization — discovery is typically guided selling or rep-led, not a native buyer-conversation core. | Both, narrow seam |
|---|
| Primary user persona | Buyer self-serve (full / guided / sales-assisted); partner portals; internal presales paste-in. — Escalation target: < 10% of sessions. | Rep, deal desk, and pricing operations users in Smart CPQ by default. — Headless partner portal and B2B eCommerce are marketing themes — often SI-built on APIs with Conga as rules core; turnkey buyer depth versus custom portal is program-dependent. | Depends on your gap |
|---|
| Deployment model | Standalone without a Conga Smart CPQ tenant: chat widget, iframe, JS snippet, embedded page, headless API, messengers. Connects to any CRM or ERP stack. | Cloud SaaS on the Conga revenue stack; PROS B2B absorbed 2026-02-02. — Requires Smart CPQ subscription, catalog modeling, and post-close packaging confirmation with Conga. | Talkulate AI CPQ job |
|---|
| PROS B2B lineage & M&A scope | No dependency on the Feb 2026 PROS B2B close — deploys as a buyer surface while Conga SKU maps and license portability are clarified. | PROS Holdings B2B business closed into Conga on 2026-02-02; PROS Travel excluded. — Public CPQ URLs now brand Conga Smart CPQ. Orderable SKU map for legacy PROS customers is not published in our review. Not established in PROS / Conga Smart CPQ public CPQ materials reviewed (May 2026); verify with PROS / Conga Smart CPQ. | Depends on your gap |
|---|
| Entry UX | Plain-language task entry with optional persona routing (B2B technical / B2C / installer / fleet). — Contextual goals, not rigid decision trees. | Guided selling in Smart CPQ UI. Winter 2026 Selling Agent adds rep-side natural-language product find and add-to-quote — buyer NL entry remains a separate build or front door. | Talkulate AI CPQ job |
|---|
| Real-time option swaps | Buyer swaps a compatible alternative → Engineer Agent revalidates the full BOM and reprices instantly. Out-of-rule swaps are blocked at the engine level. | Constraints-based logic and formula engine support repricing when rules and UI latency allow; sub-second performance at 10,000+ lines is vendor marketing — tenant benchmarks are buyer-verify. | Talkulate AI CPQ job |
|---|
| Self-serve modes | Three built-in modes: full, guided, and sales-assisted; benchmark < 10% escalation. | Headless partner portal and B2B eCommerce are positioning themes — API-first with SI or partner build is common; coexistence with a productized buyer front door such as Talkulate AI CPQ is a typical pattern. | Both, narrow seam |
|---|
| Discovery & pricing |
| Discovery model | Interviewer Agent: contextual goals — 6–10 questions, 4–8 minutes to validated BOM in reference case. | Guided selling and rep-led discovery in Smart CPQ. Selling Agent (Winter 2026) is rep-side NL catalog search — not documented as buyer configured-line constraint proof. Not established in PROS / Conga Smart CPQ public CPQ materials reviewed (May 2026); verify with PROS / Conga Smart CPQ. | Talkulate AI CPQ job |
|---|
| CPQ workflow coverage | Buyer-facing CPQ slice: discovery through validated configuration, pricing, and proposal — governed quote record may live elsewhere. Entry through handoff (CRM, ERP, or downstream Smart CPQ) in one buyer engine when Talkulate AI CPQ-only or as the front layer in coexistence. | Configure, price, quote in Smart CPQ; document outputs; one-way CRM sync; POM price recommendations in quoting when modules are licensed — CLM depth is suite-wide, not fully mapped on this Smart CPQ slice. | Both, narrow seam |
|---|
| Pricing engine depth | Volume tiers, bundles, regional lists, SLA adders, contract rebates, promotions, multi-currency — validated in the same Engineer Agent pass. | Commercial rules in CPQ plus Smart Price Management price books, scheduled generation, rebate spreadsheet upload, and POM optimization workflows when pricing-science modules are in scope. | PROS / Conga Smart CPQ job |
|---|
| Pricing science & rebates (POM) | Commercial rules live in the configurator database — integrates to downstream price optimization stacks via scoped hooks, does not ship a parallel POM SKU. | Conga Price Optimization and Management (POM) modules — price optimization, price management, rebate management — marketed with API ties to CRM, ERP, CPQ, and eCommerce. Neural-network optimization examples are vendor-scoped, not generalized proof. | PROS / Conga Smart CPQ job |
|---|
| Output formats | Interactive commercial proposal + PDF + per-line reasoning + compatible swaps + branded PDF. | Quote and proposal outputs in Smart CPQ; price-driver transparency for reps on historical deal guidance — buyer-facing per-line reasoning log is not established in PROS-primary CPQ docs reviewed here. | Talkulate AI CPQ job |
|---|
| Approval & discount governance | Discount threshold approvals available as add-on — often optional on buyer self-serve where SKU compatibility and price accuracy are the bottleneck. | Historical deal guidance with target price and discount ceiling suggestions; POM-linked recommendations — mature rep pricing-governance adjacency when modules are licensed; buyer paths may defer until revenue risk requires them. | Depends on your gap |
|---|
| System UI languages | Multi-language requires scoped retraining per language for conversation flows. | Winter 2026 documents eight OOTB system UI languages (English, French, Spanish, German, Italian, Portuguese, Chinese, Polish) — vendor-maintained for system UI and error messages. | Depends on your gap |
|---|
| Validation |
| Validation method | Deterministic Engineer Agent against live PostgreSQL via MCP — no document retrieval for core compatibility. | Constraints-based configuration engine (marketing) plus formula engine and quote multi-selection data types (Winter 2026 release notes) — rules-driven, not LLM-guesswork for core quoting. | Both, narrow seam |
|---|
| Invalid configuration handling | Engineer Agent blocks incompatible combinations with per-line reasoning before handoff. — In one published hardware reference (~3,400 SKUs), standard BOMs no longer required a separate engineering review after catalog-backed validation. | Constraint and formula logic block disallowed combinations when rules are complete; gaps are admin modeling work. Selling Agent search does not replace constraint proof for buyer sessions. | Depends on your gap |
|---|
| Winter 2026 AI agents | LLM only in Interviewer Agent; Engineer Agent is non-generative for configuration choices. | Pricing Agent (anomaly alerts), Selling Agent (NL product find on quotes), Operations Agent (data-loading troubleshooting) — rep and ops surfaces with chat history and credit consumption tracking; GA and data residency are buyer-verify. | Both, narrow seam |
|---|
| Dual-agent architecture | Interviewer and Engineer fully separated; Engineer only runs allowlisted DB queries. | Smart CPQ core is rules- and constraint-driven — not LLM-fronted for configuration. Winter agents are additive assistance layers. | Talkulate AI CPQ job |
|---|
| Catalog source connectors | PostgreSQL, MySQL, MSSQL, SAP, NetSuite, Dynamics, Salesforce CPQ, Excel, PDF specs, XML, REST, file drops. | Smart CPQ catalog plus one-way CRM sync to Salesforce or Dynamics (Winter 2026); API-first ERP posture in marketing — SAP S/4HANA and Dynamics F&O demo claims need production connector proof. | Depends on your gap |
|---|
| Per-line audit trail | Per-line reasoning: which constraint triggered each selection — buyer and rep facing. | Price-driver transparency for reps on historical guidance — buyer-facing configured-line reasoning comparable to Talkulate AI CPQ is not documented in PROS-primary CPQ sources. | Talkulate AI CPQ job |
|---|
| Observability & tracing | Langfuse per session (Interviewer and Engineer traces separately); rate limits on chat and overview endpoints. | Agent credit consumption tracking noted in Winter 2026 release notes; LLM trace layer for buyer sessions not established in public Smart CPQ materials reviewed here. | Talkulate AI CPQ job |
|---|
| Channels & time to value |
| Embed channels | Widget, iframe (~10 min deploy), JS snippet, embedded page, headless API, internal-tool mode — productized buyer shell without building on Smart CPQ APIs first. | API-first Smart CPQ with headless partner and eCommerce positioning — SI or partner builds the external shell on Conga rules; many programs pair a custom or Talkulate AI CPQ-class front door with Smart CPQ as quote engine. | Both, narrow seam |
|---|
| Async messenger channels | WhatsApp, Telegram, Teams, Slack, Messenger, LinkedIn — same validation engine, scoped per channel. | Messenger-native CPQ intake is not documented as a standard Smart CPQ feature. — Not established in PROS / Conga Smart CPQ public CPQ materials reviewed (May 2026); verify with PROS / Conga Smart CPQ. | Talkulate AI CPQ job |
|---|
| Time to production buyer surface | Envelope 5 days–6 weeks; typical 3–5 weeks; reference case 5 weeks. | Smart CPQ modeling plus optional POM, rebate, and price-book tracks — calendar lengthens with pricing-science scope and post-close SKU clarity. | Talkulate AI CPQ job |
|---|
| Customer effort | 10–15 hours over 2–3 weeks (vendor claim for standard deployment). | Heavy catalog modeling, CRM sync mapping, pricing-science parallel workstreams, and M&A packaging reviews — effort scales with module scope and SI availability. | Talkulate AI CPQ job |
|---|
| "No structured catalog" path | Data structuring service ($3,450–$17,250 one-time) for Excel, PDF, and tribal knowledge. | Catalog and constraint modeling are admin- and partner-scoped; no public fixed-fee structuring SKU on Smart CPQ pages reviewed here. | Depends on your gap |
|---|
| Verticals | Eight first-class vertical packs on one engine. | Complex B2B and manufacturing positioning on Smart CPQ surfaces — partner-led verticalization, not eight named Talkulate AI CPQ-style packs. | Talkulate AI CPQ job |
|---|
| Security & governance |
| Error mode | Refuse: incompatible combinations are blocked before the buyer accepts output; the engine does not publish out-of-rule builds from document similarity. | Constraint engine blocks disallowed combinations when modeled; incomplete rules can produce plausible but incorrect quotes — remediation is modeling, not swapping the CPQ UI. | Depends on your gap |
|---|
| Regulatory constraint encoding | CE, FDA, UL, NSF, RoHS, and similar rules encodable in validation engine. | Implementation-dependent in constraint and formula rules — depth depends on SI and customer team. | Depends on your gap |
|---|
| Anti-prompt-injection | Engineer Agent allowlisted queries only; vendor does not need internal margin data. | Smart CPQ core quoting is not LLM-fronted for configuration; Winter agents are separate surfaces — confirm agent data posture in DPA. | Talkulate AI CPQ job |
|---|
| GDPR / PII posture | DPA; no training on client catalog or conversations; data minimization. | PROS B2B–specific SOC artifacts post-close versus Conga Subscription Services legal exhibit were not mapped in this publication pass — verify with Conga. | Depends on your gap |
|---|
| Multi-tenant isolation & hosting | Cloud AWS/Azure EU regions; per-tenant DB. — On-prem: $69,000 enterprise license. | SaaS posture on Conga revenue stack; PROS-specific residency and isolation details post-close are buyer-verify. | Depends on your gap |
|---|
| Built-in analytics | Conversation funnel analytics and demand-sensing export. | Conversation funnel analytics for buyer sessions are not established in Smart CPQ-primary docs reviewed here. | Talkulate AI CPQ job |
|---|
| Outcomes & commercial |
| Quote cycle (reference metrics) | See the reference deployment band below. | Vendor customer stories on Smart CPQ pages are marketing — no matched public study on equivalent buyer self-serve catalog complexity in this pass. | Talkulate AI CPQ job |
|---|
| First-pass accuracy | See the reference deployment band below. | No matched public study on equivalent catalog complexity in this publication pass. | Talkulate AI CPQ job |
|---|
| Quote capacity | See the reference deployment band below. | Buyer-facing self-serve capacity uplift is not published in Smart CPQ materials reviewed here. — Not established in PROS / Conga Smart CPQ public CPQ materials reviewed (May 2026); verify with PROS / Conga Smart CPQ. | Talkulate AI CPQ job |
|---|
| Conversion uplift | A separate buyer-facing pilot (not the server-reseller reference case) reported better web conversion and fewer “waiting for a quote” drop-offs versus that pilot’s own baseline — treat as directional; measure on your traffic. | Conversion uplift from external Smart CPQ UX is not published in marketing materials reviewed here. — Not established in PROS / Conga Smart CPQ public CPQ materials reviewed (May 2026); verify with PROS / Conga Smart CPQ. | Talkulate AI CPQ job |
|---|
| RFQ unit economics | Illustrative unit economics only (not a price quote for your tenant): manual complex RFQs are often cited around $230–$460 versus automated self-serve often cited up to about $12 per RFQ at volume — use Talkulate AI CPQ pricing and ROI tools for your case. | No public per-seat list grid for Smart CPQ or POM in our review — contact Conga for a quote. Stale aggregator bands are not used as fact on this page. | Depends on your gap |
|---|
| Pricing model | Implementation ($18,400) + monthly ($1,725, 600 dialogs) + per-dialog overage + integration ($115 / hour). Enterprise on-prem: $69,000. | PROS / Conga Smart CPQ and POM: contact sales — no public list grid located in our May 2026 review. Packaging merges with Conga bundles post-close; net TCO requires AE validation. | Depends on your gap |
|---|