SquadStack

SquadStack positions an India-focused voice-AI system for revenue calls. It advertises 50 lakh-plus AI calls daily, 90% lead connectivity, 40% more conversions and 3x lower CAC, and lists sales/support workflows across BFSI, education, logistics, ecommerce, healthcare and other sectors.

The live stack conducts Indian-language sales, qualification, booking, collection, support and onboarding calls; extracts outcomes; handles interruptions and language switching; and can hand a live call to a human. The page says it handles 400,000-plus calls per day and combines AI and human-in-the-loop quality analysis.

The sales solution qualifies, follows up and converts leads through voice, WhatsApp, SMS, email and in-app touchpoints; writes notes, dispositions and next steps to CRM; automates retries and reschedules; supports live transfer; and covers ecommerce cart recovery, assisted purchase, upsell, repeat purchase and logistics-adjacent workforce use cases.

The support product handles inbound voice, multi-intent queries, order/delivery status, billing, payments, returns, refunds, cancellations, tickets, onboarding and KYC documents. It preserves cross-channel context, scales inbound calls, and hands unresolved cases to humans with history and extracted data. Named integrations include Zendesk, Freshdesk, Salesforce Service Cloud, Zoho Desk and LeadSquared, plus custom APIs.

The workforce solution calls, screens, schedules, follows up, guides documents/KYC/training, and activates delivery riders and gig workers. It connects to ATS, HRMS and internal tools through open APIs, supports Indian languages, and can transfer candidates to live recruiters. The page states one-week go-live for this use case.

The collections solution captures promise-to-pay amount and date, triggers reminders, detects broken promises, orchestrates calls with WhatsApp/SMS/payment reminders, applies bucket-specific rules, and escalates disputes or negotiations. It integrates through APIs with loan-management, collections, CRM and payment systems; scripts, retries and escalation thresholds are configurable.

Lead data enters from customer CRMs or manual imports, is deduplicated and checked for DND, prioritized through attributes and FIFO/LIFO rules, and routed through configurable calling, email, SMS, WhatsApp and IVR steps. The system A/B tests contact timing/channel and automatically checks connectivity discrepancies. Skills-based matching and fraud controls apply to human telecallers.

Campaign dashboards expose leads, calls, connectivity, quality, turnaround, conversions, cost of sale, lead journey, attribution, funnels and agent performance. SquadStack says it captures all touchpoints and call recordings and maintains an AI-assisted quality audit trail.

The quality product combines AI scoring and manual audit, publishes smart-sampling rates of 12% and 15% in different sections, uses 23 configurable parameters, monitors drift, and provides quality scores, call categories, transcripts and filters. These are quality controls, not proof that every call is manually audited.

SquadStack discloses per-minute billing for live connected AI-call time with no platform or seat fee. Voice AI, trained human-advisor handoff pods, lead management, experimentation, quality audits, two-way CRM sync, compliance support and a dedicated customer-success manager are bundled. Pilot, Scale and Strategic engagement shapes are public, but no rupee rate, minimum volume or contract value is disclosed.

The in-app assistant provides real-time voice guidance for sales, onboarding and support on web, Android and iOS. The page lists Indian data residency, role-based access with audit trails, PII redaction, encryption, ISO 27001 and SOC 2 Type II claims.

Persistent Memory stores configured lead-level facts across calls using a masked identifier. Customers define up to 20 extraction points per agent and can use refresh or rolling-history modes plus verbatim or AI-summarized context before a follow-up call.

Lift analyzes production calls, proposes small instruction changes, tests variants on live traffic and increases traffic only when conversion data supports them. A human reviews every proposed production change, and compliance, pricing and identity sections can be locked. The post reports anonymous production tests, including appointment booking rising from 1.5% to 2.7% across more than 12,000 leads and branch visits from 2.2% to 4.3%.

Optimize runs controlled tests on voice, prompts, workflow, channel and cadence, tracks conversion/connectivity/qualification/cost in real time, flags statistical significance and supports gradual traffic ramping. It is built into the platform and connects to quality audits, lead scoring and omnichannel execution.

Overwatch is live for SquadStack's internal campaign team. It traces end-to-end, STT, TTS, model and tool-call latency; tool failures; telephony; DND/DNC; stuck leads; errors; and affected call IDs. Automated buyer-impact and pitch checks are described as rolling out campaign by campaign, so that layer is partial.

SquadStack says its context-management layer maintains long-call history and its RAG layer retrieves product, policy, pricing and FAQ content in under 100 ms. It reports about 97% retrieval accuracy in blind tests including an 87-page manual; the evaluation method and independent validation are not public.

An inline processor masks Indian PII between speech-to-text and downstream storage/model layers, uses a six-turn window to catch split values, isolates redactors per call and maintains a 400-plus-case regression suite. SquadStack says Angel One's security and compliance teams approved the production deployment; no customer-owned corroboration was retained.

Terms identify Squadrun Solutions Private Limited and affiliates as SquadStack and define services to include telecalling, lead qualification, sales, analytics, campaign management, dashboards and communication interfaces. They require customer consent for voice capture/use, state AI output can be inaccurate, and require human review and backup controls for critical or legally significant decisions.

The policy says Squadrun Solutions Pvt Ltd is a subsidiary of SquadRun Inc., describes a distributed-worker application and related personal-data processing, states India/New Delhi governing jurisdiction, and links an official trust portal. Some description still reflects the older worker-network product, creating a legal/product-currentness ambiguity.

The certificate covers Squadrun Solutions Pvt Ltd's AI-services BPO/outbound-calling system, was initially issued 11 November 2024, lists surveillance validity to 10 November 2025 and recertification on 10 November 2027. The artifact does not establish that the missed surveillance checkpoint was completed.

SquadStack describes its vision as blending AI precision and human empathy across enterprise customer conversations. It reports 200-plus SquadStack members and 10,000-plus sales experts; the two populations are distinct and neither is a regulator-verified employee count.

The company names Apurv Agrawal (CEO), Kanika Jain (Product), Vikas Gulati (Technology) and Rishabh Ladha (Business) as co-founders. It lists BII, Chiratae and Blume as investors and reports ₹190 crore raised across Series A and B.

SquadStack says a ₹140 crore Series B was led by Bertelsmann India Investments with Chiratae and Blume participating. The article describes a SaaS-enabled sales-talent marketplace, 170,000 leads managed daily and 3x year-on-year company growth, but discloses no revenue amount. Its displayed 2026 publication date appears inconsistent with historical framing, so the round date is not treated as resolved.

The named vendor case reports 70% connectivity and 2x conversions. The body says Amity's 30–40% internal connectivity nearly doubled in two weeks and that a dedicated human calling team contacted all leads with omnichannel support.

The named case headline reports 87% connectivity, 46,000-plus demos booked and under-five-minute turnaround. The body describes a fully managed, trained and monitored human telecalling team rather than autonomous AI voice execution.

The named case reports 90% connectivity, 30% more conversions and a 30% boost in water-purifier AMC sales. Its body separately says connectivity improved 50%, from a starting operation of over 55 lakh leads per month and 3–5% conversion. The deployed system combined managed callers, AI knowledge/quality tools and outreach workflows.

The anonymous case reports 90% connectivity, 30% higher appointment bookings and 67% fewer attempts per lead. The body says connectivity rose from 65% to 90% with 2–3 attempts instead of 5–6, while also describing 25% first-attempt connectivity and 6.5% appointment conversion as different pre-deployment baselines.

The named case headline reports 86% connectivity, 50% lower CAC and an 82% conversion boost; the title frames the 82% as loan-disbursal growth. The body describes managed callers, AI knowledge management, outreach, reporting and quality monitoring but does not reconcile the headline outcomes.

The named case reports 77% connectivity, 35% tests rescheduled and 15 lakh leads reached in one week through an on-demand human telecalling team using calls and WhatsApp. It is evidence of productized managed execution, not autonomous Voice AI.

The anonymous case covers outbound buyer qualification and inbound missed-call recovery. It reports 70% better reach, 50% stronger qualification, 24% more complete lead data, 92% inbound connectivity and 45% lower cost per qualified lead, but its key-outcomes section instead says 40% higher connectivity and 50% higher conversion.

The anonymous case reports 3x higher conversions and 3.2x lower average handling time. The autonomous Voice AI performs outreach, qualification and follow-ups and writes CRM updates in real time; detailed baseline values and measurement window are absent.

The named managed-sales case says Choice tested nearly 5x its normal lead volume, averaged 72% connectivity in Q1 2022 and reached under ₹90 CAC in March 2022. The title claims a 70% CAC reduction, and the headline lists 10-plus campaigns, without a disclosed comparison formula.

The anonymous managed-caller case headlines 87% connectivity, 8x ROI and 11% average conversions. Its body says prior abandoned-cart conversion was 20–25%, conversion became 2x organic conversion and more than 55% of leads converted within 48 hours. The scopes cannot be reconciled from the page.

The named AI-versus-human case reports onboarding about 3x faster, AI connectivity of about 72% versus 62% for humans, qualification improving from 7% to about 17% versus a stable 14–15% human rate, and nearly 4x lower acquisition cost per qualified rider.

The named vendor case reports more than 100,000 autonomous buyer conversations per day, 75%-plus connectivity versus 50–60% for humans, 20% higher conversion on the same lead pool, 15% lower cost per confirmed lead and 95% outcome accuracy. It says the system moved from pilot to full production in three months and hands qualified outcomes into the commerce workflow.

The anonymous case separates sales outreach from listing-data verification. Sales results include 85% connectivity, 25% qualification, 70% live transfer in an earlier version and 2.5% direct appointment booking currently. Verification reports 91% connectivity, about 30% higher connectivity and 55% higher complete-data confirmation than humans, 95%-plus capture accuracy, about 60% lower cost and about 5x volume scaling.

The anonymous human managed-service case reports up to 65% improvement in internal-team productivity, 50% downward and 200% upward weekly scaling, 72-hour go-live, 145-plus campaigns, 63-lakh-plus processed leads and 30–40% conversion improvement. In December 2021 it says 35 campaigns processed more than 13 lakh leads; the headline separately states 86% connectivity.

The anonymous insurer case reports 85% connectivity, 25–30% above humans, about 33% lower AHT, unchanged conversion, 50–60% lower operating cost and more than 10 lakh leads handled monthly. It reports removal of a 24–48-hour response delay, automated QA with 3% human sampling, CRM sync, segmentation, callbacks and warm transfer. The headline calls the 60% result lower CAC while the body calls it operating cost.

The named case headline reports 85.5% connectivity and 25% more appointments booked. The body describes a flexible fully managed, trained and monitored human telecalling team with usage-based productive-connected-minute pricing.

The named managed-calling case headlines 89% connectivity and 40% more loan applications. The body reports 25–40% outperformance of internal submit-rate benchmarks and a 154% connectivity improvement through system-driven guided calling. A quoted 10x rise in Moneyview disbursals over two years is company context, not attributed to SquadStack.

The anonymous case explicitly separates AI ownership of high-volume worker onboarding/milestones from human handling of complex conversations and exceptions. It reports 90%-plus connectivity, 40% lower workforce-acquisition cost, about 15% of onboarded workers completing first login/first order and about 10% reaching the tenth order.

The named human managed-service case headlines 75% connectivity, 50% lower survey cost and 15% successful surveys. Its body describes productive-connected-minute pricing, flexible work-from-home callers, 12% call sampling with up to 3% manual review across 23 parameters, and omnichannel after-hours coverage.

The named case says SquadStack's Humanoid AI automated Shiprocket presales calling, classified buyers versus sellers and drove sellers toward first-time recharge. It reports 3-lakh-plus leads, 4x engagement scale in one quarter, 5x seller-identification accuracy, 30–35% lower inbound-team workload and 5x first-time recharge; a headline 9x connectivity claim lacks a body baseline.

The named case headlines 55% AI containment, 70% cost reduction and 86% positive CSAT. Its body reports 79–81% AI CSAT, 83–84% human CSAT and 84–85% hybrid CSAT, with 2x greater feedback propensity after AI calls. The headline and detailed CSAT values conflict; a planned 60–70% AI share is future, not achieved.

The named managed-human case headlines 77% connectivity and 36% successful surveys. It says an on-demand caller team ran seven campaigns and reached more than 29,000 vendors, with all calls recorded and AI used for voice-of-customer analysis.

The named managed-sales case reports 75% connectivity, 119-plus campaigns and 2-crore-plus processed leads. The title says account activations increased 40%, while the body says the activation rate reached 40%; it also reports 55% activation on a 41-expert tailored campaign and more than 740 callers assisting account opening in the latest month.

The named case reports about 87% higher usable-rate-card collection versus an earlier below-20% success rate, about 40% lower AHT, about 85% connectivity and a headline 50% lower cost. The Hindi/Hinglish Voice AI was trained on 13,000 tagged leads and captured route-, vehicle- and return-specific pricing.

The current sitemap enumerates 24 distinct URLs under /case-study/, all retained as separate normalized cases. It also exposes the current product, pricing, legal and product-announcement surfaces used in this dossier.