← Agent catalogue
AI Agent Profile · LendingIQ · Bengaluru
Sales Development Rep AI
Function: SDR / Telesales ExecutiveInvoked via: CRM lead queue — daily batchRuntime: AWS Bedrock · ap-south-1Model: Claude Sonnet 4Context window: 200K tokens
DivisionGTM Sales
Resume
What this agent does
The Sales Development Rep AI works the top-of-funnel lead queue — qualifying inbound and outbound leads through structured call scripts and WhatsApp follow-ups, capturing lead data and intent signals in the CRM, and routing qualified leads to the Account Executive AI for application assist and deal closure. It handles volume at the top of the funnel so human sales staff can focus on complex relationship and application-stage conversations.
Primary functions
Lead Qualification
Every lead — on first contactINVOKED WHEN: lead enters the CRM queue and is assigned to the SDR AI for first contact
- Reads the lead record — source, product interest, segment indicators, and any prior contact history — and conducts the qualification call using the approved qualification script for the applicable product and segment: confirming loan purpose, ticket size range, business vintage or employment status, existing obligations, and contact preference. The qualification call is not a credit assessment — it gathers the information needed to determine whether the lead is worth routing to an Account Executive.
- Scores the lead against the configured qualification criteria: minimum business vintage, minimum income band, product eligibility indicators, and expressed intent level. Produces a qualification outcome: Qualified (route to Account Executive AI), Nurture (follow up after defined period), Not Qualified (log reason and close). Qualification criteria are configured by the sales team — the agent applies the criteria, it does not set them.
- Does not discuss specific interest rates, make pricing commitments, or assess creditworthiness during the qualification call. If a lead asks about rates, the script provides a range with a clear statement that the exact rate depends on the credit assessment — the agent does not quote a specific rate.
Output: Qualified lead record — qualification outcome, score, key data captured (loan purpose, amount, segment indicators), and routing instruction. All contact details and conversation summary logged in CRM.
WhatsApp Follow-up
Leads not reached on first callINVOKED WHEN: outbound call is unanswered or a lead has requested WhatsApp contact as preference
- Dispatches the WhatsApp follow-up template for the applicable product and contact attempt number — a brief, friendly message confirming LendingIQ's call attempt, stating the product context, and providing a callback link or CTA. Three-attempt WhatsApp sequence before the lead is routed to the human SDR team for a different contact approach.
- Classifies WhatsApp replies: interested (route to Account Executive AI with summary), callback request (schedule and route), not interested (log and close), stop/opt-out (immediate do-not-contact log), or question requiring detail (route to Account Executive AI). Unclassifiable replies are escalated to human review — the agent does not compose novel replies to unclassified messages.
Output: WhatsApp contact log — template sent, reply received, reply classification, and routing action. Opt-out processed immediately and logged.
Hard guardrails
Will notCommit to specific interest rates, fees, or loan terms. Any rate or term discussion requires an Account Executive — human or AI.
Will notContact leads before 8:00 AM or after 7:00 PM. This constraint is hardcoded identically to the collections contact agents.
Will notContinue contact after an opt-out signal. Opt-outs are processed in the same session with immediate effect.
Known limitations
Call script effectiveness depends on the script quality, not the agent's capability.A poorly designed qualification script will produce low-quality lead qualification regardless of how well the agent executes it. Script quality is a human design responsibility — the agent is the executor, not the author. Regular A/B testing of script variants should be a standard practice.
The agent cannot read tone or emotional context on a call the way a human SDR can.A lead who gives technically qualifying answers but who is clearly unenthusiastic or confused will not be flagged as a relationship risk the way a skilled human SDR would notice. Build a 'human escalation on any uncertainty' rule into the script so the agent transfers to a human when the qualification is technically borderline rather than continuing to push for a qualification outcome.
Important Reads
Learn more about how to deploy Sales Development Rep AI to your lending workflow.