Katalyst is the AI sales agent for teams on Salesforce. Hang up a call and it's already done: notes summarized, records created, fields updated, follow-up drafted, next step set. It runs 24/7, reading every call, email, and calendar, surfacing the right signals, prepping you for each meeting, and prompting you to act: follow up here, this one’s slipping. What's new: AI Resolution on every account, meeting recorder, hygiene scores, deal patterns. Built for enterprise sales teams.
How's this really different to Agentforce? What's different about Katalyst?
Katalyst
@mukesh_pandey6 Agentforce wants you doing more inside Salesforce. Katalyst does the Salesforce work for you, so reps barely touch it. One deepens the platform, the other removes the busywork.
Katalyst
@mukesh_pandey6 To add, the tell is where each one lives. Agentforce assumes the CRM is where reps already work and layers agents on top. Katalyst assumes the opposite, the work happens in calls and emails, so it captures it there and keeps Salesforce updated for you. One makes the CRM busier, the other makes it disappear from your day.
Congrats on Shipping! Looks like a promising product here! In my experieence is the graveyard of CRM tools is full of products that made data entry slightly less painful and called it a solution. How's the auto-capturing of data working in Katalyst? The problem usually isn't the UI it's that data is scattered, async and AEs don't really log the things. Running a SaaS with a mix of inbound and long-cycle partnership deals, so the fragmented relationship problem is constant for me. How's Katalyst solving for this?
Katalyst
@jalan_kshitij Thanks! And you're describing exactly the bet we made. You're right that the graveyard is full of "faster data entry" tools, so we went the other way: capture can't depend on the rep doing anything at all. Katalyst plugs into where the conversation already lives. Connect your inbox and your notetaker, and every email and call gets matched to the right deal and contacts automatically, then turned into the CRM updates, follow-up tasks, and next steps that should come out of it. The rep's only job is a quick yes or no on what it drafted. Nothing to log, nothing to remember to log.
The long-cycle part is honestly where it earns its keep. A partnership deal goes quiet for two months, then a thread wakes up, and all the context from every email and call along the way is already sitting on the deal, not in the rep's head. New people show up mid-deal on those too, and Katalyst spots contacts it hasn't seen and offers to add them. Would love to hear how it holds up against your setup, the inbound plus long-cycle mix is a great stress test.
Does each customer end up with their own sales memory? I'm wondering whether the AI gradually learns how our company sells?
Katalyst
@aryan_mehan1 Yes!! each organization gets their own sales memory that compounds over time.
Every email, meeting, and deal your team processes teaches the system about your company's playbook. It learns what actions tend to move deals forward at each stage, which signals matter most, how your reps structure conversations. The more deals you run through Katalyst, the smarter the agent become because they're building a memory on how your company actually wins, not generic AI patterns.
The auto-drafted follow-up sounds useful but here’s my concern with every AI email sounding like AI, can reps actually tweak it or do they have to copy-paste into Gmail and rewrite from scratch?
Katalyst
@saanvi_jalan Great catch! Reps customize the drafts directly in Katalyst — no copy-paste cycle. They view the draft, edit tone/details inline, and send.
On the "AI-sounding" concern: that's exactly why the template system matters. When a rep saves past emails tagged as "contract negotiation" or "expansion play," Katalyst learns their voice — cadence, formality, length, structure. When it generates the next draft, it matches against those templates first, so the AI output inherits their actual writing style, not generic corporate boilerplate.
More templates saved = better the agent learns your team's voice. It's not "fill the template," it's "generate like this person writes."
The draft is a starting point, but the customization (the 30 seconds of tweaking) is where the rep's personality lives. That's what kills the generic AI sound.
I believe that Katalyst has a lot of products at once (reads your calls, emails, and calendar, and turns it into action) I was wondering if it makes the onboarding difficult. Which workflow do you recommend customers to try first?
Katalyst
@reda_roqai_chaoui Fair question, we thought about this a lot. Instead of dumping all the connections (calls, email, calendar) on you at signup, onboarding is guided: you walk through a short setup flow where the product, the Katalyst's own agent walks with you tells you what to connect and why, one step at a time, and you see real output from your own data as you go — not a demo, your actual first draft or first insight.
So the "lot of products" doesn't show up as a wall of settings. It shows up as a conversation: connect this, see this work, connect the next thing when you're ready. Most teams are through full setup in under minutes and already have a first AI action to react to.
If you want a specific recommendation: whichever data source has the most recent activity for your team (usually email or calls) is what surfaces value fastest, and the onboarding flow actually points you there automatically rather than making you guess.
What are the "signals" you keep mentioning? What does it mean?
Katalyst
@sanvi_tiwari Signals are the things that suggest a deal's changing, a prospect raising budget, a new stakeholder joining, activity going quiet, a renewal date approaching. Katalyst surfaces them so reps act on the moment instead of missing it.
Katalyst
@sanvi_tiwari happy to add some color here. Signals are pieces of news about the companies you're selling to: a funding round, a leadership change, an acquisition, a hiring push, a new product launch. Katalyst watches for these across your accounts so you don't have to, scores them so the ones that matter most rise to the top, and attaches each one to the account it belongs to. The point is timing. A prospect who just raised money or got a new VP is a very different conversation than one you ping out of the blue, and signals hand reps those openings while they're fresh. You pick the types of news you care about, so it stays a signal and not a feed of noise.
Is this only for tech companies or would it work for any industry?
Katalyst
@hanish_bansal Any industry where the sales team runs on Salesforce. Fintech, healthcare, manufacturing, govtech, you name it. The pain of manual CRM data entry and inconsistent pipeline management is universal across B2B sales. The AI agent works with whatever Salesforce configuration you have, regardless of industry.
Katalyst
@hanish_bansal Adding to Avneet, non-tech teams often benefit most, manufacturing and healthcare reps are in the field, not living in Salesforce, so the further you are from a desk, the more the auto-updating matters.