SquadStack.ai

Celebrating the people who've stuck around long enough to get the jokes, make the jokes, and become the jokes. Here's to another round. 🥳 P.S. Want in on the jokes? We're hiring: https://lnkd.in/g4FBQZ_r

Most AI PM roles are about roadmaps and PRDs. This one puts you in front of enterprise CXOs, owning how their voice AI system actually performs. At SquadStack.ai, we power 50 lakh+ autonomous sales calls every day for companies like IndiaMART, Bank Bazaar, Tata Digital, and more. You will manage enterprise accounts end-to-end, designing their voice AI agent, shaping call flows, and making sure the AI delivers in their specific business context. You will need enough product instinct to know what to build and enough client context to know what actually matters. 📍 Noida Apply on the link in the comments.

Celebrating Ayush Shanker as he completes 10 amazing years at SquadStack. 🚀 He has been a backbone for his team all these years, helping them with genuine feedback that's sharp and zero fluff, and then somehow make the whole thing feel like it was their idea. He’s not a manager but a mentor to his team! Someone who gives you the space to figure things out while quietly having your back the entire time. But that’s not all, there's a witty side behind all that calm demeanor! Ask anyone who's been on a Goa trip with him. Congratulations Ayush, we're lucky you never learned how to stress. 🥳

Most of these people ship voice AI together every day. Turns out they're equally intense about finding a water bottle before the other team does! At SquadStack.ai, we don't take any challenge lightly 🤭

Celebrating Manmeet Singh as he completes 10 years of being a part of our Squad! Manmeet joined before we had a product, a plan, or honestly any idea what we were doing. Through every pivot, every fire drill, every "we’ll have to figure this out by tonight", Manmeet carried the team when it got heavy and made it look easy (it wasn't). His impact is way higher than the time he spent, the "OG'iest of the OG"! But the best part is these 10 years weren’t spent with colleagues but with his closest friends for life! 10 down, many more to go 🚀 Congratulations Manmeet! We appreciate you for everything you have done and the energy you bring to the Squad!

SolarSquare just raised $53M in Series C, led by B Capital with Lightspeed, Lowercarbon, Good Capital, Zerodha, and Better Capital. 50,000+ Indian homes now run on solar. 29 cities. And a residential solar brand that figured out how to make adoption painless for homeowners. We've had a front-row seat to this for over 4 years as their sales partner, handling the conversations that get homeowners from curious to committed. Congrats Shreya Mishra, Neeraj Jain, Nikhil Nahar. We have loved working with Munish Vashishat, Rashmi Dwevedi and the SolarSquare Energy team! Sakshi Kulhans Anunay Kumar

Day 5 of 5: Introducing Inline PII Redaction SquadStack.ai's voice AI redacts customer PII in real time, before it reaches the LLM. The calls that drive the most revenue are the ones where customers speak Aadhaar numbers, PAN, bank details, OTPs out loud. In most voice AI systems, that data sits in plain text across the full stack. STT output, transcript, LLM prompt, model provider's API, logs, QA dashboard. No redaction at any step. For teams running voice AI without this guardrail, compliance blocks the deployment. And they should. The hard part: PII on voice calls doesn't arrive in one clean line. A customer spells their PAN across four turns. They switch between Hindi and English mid-sentence. Single-turn detection misses most of it. 𝗦𝗼 𝘄𝗲 𝗯𝘂𝗶𝗹𝘁 𝗶𝗻𝗹𝗶𝗻𝗲 𝗣𝗜𝗜 𝗿𝗲𝗱𝗮𝗰𝘁𝗶𝗼𝗻 - a processor between the STT engine and the LLM with a 6-turn rolling window that catches PII split across fragments. 16 types, tuned for Indian voice conversations. A regulated brokerage's compliance team reviewed this and signed off for production. So compliance stops blocking the highest-value use cases and starts approving them. Deep dive in comments.

Day 4 of 5: Introducing RAG & Context Management Your AI agent can now handle long complex calls without a human stepping in! A few weeks ago I wrote about how we fixed 20-minute long revenue calls. How the agent could remember everything the lead said and hold up! Remembering wasn't enough on its own. In some instances, the agent just couldn't reliably answer the very hard and specific product question the entire call was building toward. Not because the knowledge was missing. Every policy detail, pricing sheet, FAQ was loaded into the prompt. The prompt was just so heavy that the model skimmed past the exact answer the caller needed. So we moved product knowledge out of the prompt entirely, and put it in a knowledge base. The agent reads from it,  only when the caller asks. Retrieval takes under 100 milliseconds. On blind tests, including an 87-page product manual, it finds the right answer 97% of the time. Context Management gave the agent memory. The knowledge base gives it depth. Together, the 20-minute insurance call, the loan advisory call with fifteen branching questions, now close without a hiccup or the need of transferring to a human.

Day 3 of 5: Introducing Overwatch A question I hear from leaders running voice AI at scale: how do we know what's actually happening on the calls? The most expensive problems never throw an error. If the greeting has a two-second lag, or the agent fumbles a rebuttal on a specific objection, or it marks a lead as qualified without capturing the right entity. You find out days later, when the numbers slip. That's why we built Overwatch. A call slows down? Overwatch tells us whether it was the STT, TTS, LLM, or a tool call. A campaign stops connecting? We see the telephony vendor, the drop reason and the DND status before anyone files a ticket. A metric spikes? You go straight from the number to the exact calls it touched, by ID, ready to open and listen to. Every problem you miss on the first call repeats on the next thousand. At ten thousand calls a day, seeing it early is the difference between scaling and quietly leaking revenue. Overwatch is now live inside our Humanoid Voice AI Agent. Full story in the comments.

Day 2 of 5: Introducing Vernacular Voice AI Agents Every voice AI platform in India lists a dozen vernacular languages. Deploying them in production is a different problem. Today we're launching vernacular voice AI agents across India's regional languages. Live in enterprise campaigns across travel, lending, and marketplace verticals. The naturalness behind them is 80% hand-built dialogue engineering: speech artifacts mined from real agent calls, per-voice filler testing on 8kHz telephonic lines, and a kill list of phrases that instantly signal AI. How we built it: https://lnkd.in/g_CZ7XED