Results
The challenge
A Bangalore-based B2B SaaS company (RevOps-adjacent, Series B, ~120 people) was drowning in inbound. The top-of-funnel was actually working — targeted content and LinkedIn outreach were driving 200+ form fills per month. The problem was what happened after the form fill.
Manual qualification took 2–6 hours depending on SDR capacity. Many leads went cold before an SDR could reach them. Around 15% were getting qualified fast enough to convert. The rest either got a generic drip email (high unsubscribe rate) or sat in the CRM untouched for days.
Leadership's first instinct was to hire 3 more SDRs. That would cost ~₹60L/year plus onboarding time. They asked whether AI could fill the gap instead.
What we built
Arthat built a sales pipeline agent that sits between the form fill and the SDR's inbox. Here's how it works:
60-second response: Within 60 seconds of a form submission, the agent enriches the lead using Clearbit + LinkedIn data, scores them against the ICP, and takes one of three actions:
- Hot lead → auto-book: If the lead scores above the threshold AND matches a decision-maker persona, the agent sends a tailored email with a Calendly link. An SDR gets Slack-notified and a CRM task.
- Warm lead → nurture: If the lead is a good fit but not a decision-maker (or signals early-stage interest), the agent enrolls them in a 4-week personalized nurture sequence. Each email is written per-lead based on their role, industry, and the content they originally engaged with.
- Not-a-fit → polite decline: If the lead clearly doesn't match ICP (wrong company size, wrong region, student/agency, etc.), the agent sends a polite decline and removes them from active pipeline. No SDR time wasted.
CRM hygiene: The agent writes clean, complete CRM notes on every lead, tags them, assigns ownership, and adds contextual notes SDRs can use on outbound. No more "who's this lead" moments.
Sequence personalization: For the warm-lead nurture, each email is generated against the specific lead's profile. An SDR can override or approve before send for any lead over a certain deal-size threshold — human in the loop where the money is.
Weekly pipeline review: A Monday-morning agent summary lands in the RevOps Slack channel: qualification rate, conversion rate, velocity, cohort performance. Things leadership used to ask for in quarterly BI runs are now real-time.
Technical architecture: HubSpot as the CRM of record, Clearbit for enrichment, Claude 3.5 Sonnet for reasoning, GPT-4o-mini for high-volume personalization, Gmail API for email send, Slack for internal notifications, Calendly for booking. Everything runs in the client's AWS Mumbai region.
Arthat AI
Want results like these at your company?
We built this for them. We can build something similar — or something completely different — for you. Start with a 30-minute discovery call and we'll tell you honestly whether AI is the right tool for your situation.
