Vedika and Ved — The Two AI Agents Inside Cruise™

Two AI Agents. One Shared Intelligence Layer.

Cruise™ is powered by two AI agents working in concert on every exception — Vedika, who manages communication, and Ved, who manages intelligence. They don't work sequentially. They work simultaneously on the same data, each doing what they are best at, so every exception gets both the right action and the right context.

Vedika — AI Logistics Communication Agent

What Vedika Does

Vedika manages all real-time communication across the logistics network. The moment Cruise confirms an exception, Vedika calls the relevant stakeholder — driver, transporter, hub team, plant team, or client — in their preferred regional language. She conducts a structured conversation designed for the specific exception type, captures the reason, confirms the commitment time, and transcribes every word into the exception ticket automatically.

If there is no answer, Vedika retries. If the issue remains unresolved after defined retry attempts, she hands the case to the next escalation level — automatically, with full context carried forward.

What Vedika Handles

• AI-powered driver calling on exception detection
• Transporter follow-up and escalation communication
• Halt verification — confirming reason for stopped vehicle
• ETA delay confirmation with revised commitment time
• Route deviation check — planned diversion or unplanned deviation
• Tracking gap follow-up — attempting to restore visibility
• Hub dwell follow-up — identifying cause of extended facility dwell
• Delivery delay update to consignee or client team
• Response capture and verbatim transcription
• Remarks and commitment logging into exception ticket
• Escalation handoff when cases remain unresolved

Vedika in Action

A vehicle on the Mumbai–Pune corridor has been halted for 45 minutes beyond threshold. Cruise detects the exception and classifies it P2. Vedika calls the driver in Marathi. The driver explains there is a tyre puncture — mechanic en route, estimated delay 90 minutes. Vedika confirms the commitment time, transcribes the response, updates the trip ETA, logs the remark to the exception ticket, and notifies the consignee with the revised delivery window. The coordinator sees a resolved ticket with full context — without making a single call.

Languages Vedika Speaks

• Hindi
• Marathi
• Tamil
• Telugu
• Kannada
• Bhojpuri
• Gujarati

Vedika identifies the preferred language from the driver profile and switches automatically — ensuring natural, unambiguous communication in every interaction.

Ved — AI Logistics Intelligence Agent

What Ved Does

Ved monitors operational patterns, identifies risks, prioritises exceptions, and supports logistics teams with structured insights and next-best-action recommendations. He is the intelligence layer that ensures the right exceptions get the right attention — and that patterns invisible to any individual coordinator become visible to the entire operations team.

What Ved Handles

• Exception prioritisation — P1, P2, P3 classification by severity and SLA impact
• ETA breach risk detection — flagging trips likely to miss SLA before the breach occurs
• SLA risk analysis — scoring active trips by probability of SLA failure
• Route deviation classification — planned vs unplanned, severity scoring
• Halt severity scoring — rest stop vs suspicious halt vs breakdown indicator
• Suspicious pattern detection — multiple halts on same corridor, unusual timing
• Transporter performance analysis — exception frequency, response rate, resolution time
• Route-level issue analysis — lanes with recurring delay patterns
• Repeated delay identification — systemic issues by transporter and corridor
• Exception summaries — structured summaries for shift handover and management review
• Next-best-action recommendations — escalate, reroute, notify consignee, flag for review
• Executive-level reporting — KPI trends, bottleneck identification, cost and efficiency impact

Ved in Action

It is 11pm. Eight vehicles from three different transporters on the Bengaluru–Chennai corridor have logged ETA breach risk in the last two hours. A coordinator managing 600 active trips would see eight individual alerts. Ved identifies the cluster: same corridor, similar timing, multiple transporters — likely a road or weather event. He surfaces a single corridor-level alert with a recommendation to check NH-48 conditions, issue a proactive update to Chennai-end consignees, and hold downstream hub scheduling until ETAs are confirmed. One action instead of eight.

How Vedika and Ved Work Together

Vedika and Ved operate on the same intelligence layer. Every call Vedika makes, every response she captures, every exception she resolves — feeds Ved's pattern detection. Every pattern Ved identifies, every risk score he generates — shapes how Vedika prioritises her next communication.

This creates a feedback loop that makes Cruise smarter with every trip:

• Vedika captures what is happening on the ground
• Ved identifies what needs attention next
• Vedika acts on Ved's priority signals
• Ved learns from Vedika's resolution outcomes

Over time, the intelligence layer builds institutional memory: recurring transporter issues, high-risk halt locations, corridors with predictable delays, hub dwell patterns by shift. This knowledge stays in the system regardless of team turnover.

AI and Human — How They Divide the Work