Why This Ex-Uber, Meta Exec Believes AI Agents Don't Have to Replace Sales Reps

Why This Ex-Uber, Meta Exec Believes AI Agents Don't Have to Replace Sales Reps

As enterprises embrace AI agents, Proshort is betting the bigger opportunity lies in augmenting sales teams with real-time intelligence, helping representatives close deals faster.

As enterprises race to deploy AI agents across business functions, Proshort is taking a different approach. Instead of building AI to replace sales teams, the startup, with headquarters in California, US, and a development hub in Bengaluru, is developing an intelligence layer that works alongside sales representatives, providing real-time guidance during customer conversations and automating repetitive work behind the scenes.

Founded by former Uber, Meta, Yahoo, and ShareChat executive Gaurav Mishra, Proshort positions itself as an AI sales platform that sits on top of an organisation's customer relationship management (CRM), sales conversations, and marketing content. The platform analyses these data sources to recommend the next best actions for sales representatives—from drafting follow-up emails and identifying the right executive to involve in a deal, to surfacing relevant marketing collateral during customer interactions.

The startup's latest focus is live AI assistance during sales meetings.

"We're bringing that actionability live into a meeting," the founder and CEO tells AIM, explaining that the system can function as an on-screen assistant, teleprompter, or even an AI avatar depending on customer preference. The objective is to replicate the practices of an organisation's best-performing salespeople and make those insights available to the rest of the team in real time.

Building on Existing Models

Proshort hasn’t built a proprietary LLM and instead relies on foundation models from providers such as OpenAI, Gemini, and Anthropic, alongside open-source models deployed in hosted environments. The engineering effort, according to Mishra, lies in identifying the most accurate and cost-effective model for each workflow.

The entrepreneur, who founded the company in 2022, says this architecture allows it to optimise both performance and inference costs while assuring customers that their enterprise data is not used to train models.

Solving Sales Productivity

The idea for Proshort emerged from Mishra’s experience building AI and machine learning infrastructure at consumer internet companies.

While platforms such as Uber and Meta invested heavily in AI-driven personalisation and recommendation systems, he observed that enterprise software—particularly sales workflows—had not seen a similar level of intelligence integration despite significant spending on CRM and sales tools.

He argues that sales remains one of the most inefficient functions in large enterprises, with high employee turnover and a majority of representatives failing to meet their annual quotas. AI, he believes, can bridge this gap by making institutional knowledge accessible to every salesperson rather than limiting it to top performers.

Gartner found that AI tools save sellers an average of 4.8 hours per week, but 72% of sales organisations fail to reinvest those savings into high-value activities. Meanwhile, 31% of chief sales officers cite difficulty proving AI ROI as a top challenge for 2026.

According to Mishra, customers using ProShort have reported up to a 25% improvement in sales win rates while reducing administrative work by 8–12 hours per representative each week through AI-generated emails, CRM automation, task recommendations, and coaching workflows.

Human-Centric AI

Even as enterprise software vendors increasingly market autonomous AI agents capable of replacing knowledge workers, Proshort is deliberately positioning its technology as augmentation rather than automation.

Mishra says one of the biggest challenges the company faces is overcoming customer concerns that AI could eliminate sales jobs. Instead, he argues that purchasing decisions will continue to depend on human relationships, with AI acting as a productivity layer that automates routine tasks while enabling sales teams to spend more time with customers.

The company's AI agents already automate CRM updates, sales forecasting, and report generation, while also functioning as virtual sales engineers during customer meetings when specialist staff are unavailable.

Growth Stage

Proshort primarily serves B2B SaaS companies but says its platform can be deployed across any enterprise selling complex products through multi-stage sales cycles. Its customers include Unicommerce, RateGain, and MetricStream, alongside firms in healthcare, cybersecurity, supply chain, and enterprise software.

Alongside the product expansion, ProShort is experimenting with its pricing model. The startup currently charges customers on a per-seat B2B SaaS model, but is also testing consumption-based and impact-based pricing with some early customers. “We’re still figuring out our pricing model,” Mishra said.

The startup recently raised $8 million in seed funding from investors including Neotribe Ventures, Correlation Ventures, and WestWave Capital.

Mishra says the company recently crossed $1 million in annual recurring revenue (ARR) within a year of launching its product. It has also signed multiple six-figure enterprise contracts, with average contract values increasing tenfold over the past year. The company currently has a 23-member team, including around 15–17 employees across product and engineering.

ProShort is now taking its AI assistance directly into live sales calls, where the system can provide real-time guidance to representatives through a text feed, teleprompter-style interface, UI support or even an AI avatar. The company is also exploring using its AI as a virtual sales engineer or solution architect during calls when specialist resources are unavailable. 

“We’re bringing that actionability live into a meeting,” Mishra adds.

According to Grand View Research, the Indian AI sales enablement market is projected to reach $1.4 billion in revenue by 2033, growing at a compound annual growth rate of 22.5%. Proshort works in a growing ecosystem that includes peers like Mindtickle, Paperflite, and Toolyt.