Public housing assistance is supposed to work like a safety net. In practice, the system behind that safety net still runs on outdated software and paper files at many public housing authorities. A single income calculation can require pay stubs, benefit letters, and manual math before a family finds out what they owe and what the government will cover. Bejoy Narayana points to a number that stuck with him early on: by his estimate, 26 million low income families in the US are one setback away from losing their housing.

Bejoy Narayana is the Founder, CEO, and Product Architect of Bob.AI, a platform that helps low income families find housing and helps public housing authorities manage the transactions behind that assistance, from income verification to voucher approval to landlord contracts.

In this episode of Lead with AI, Dr. Tamara Nall speaks with Bejoy about the meeting that pulled him out of enterprise AI consulting and into public housing, how Bob.AI calculates what a family owes toward rent, why the platform is built around three different user groups, and what it would take to help families move from rental assistance toward homeownership.

A Four Hour Lunch That Redirected a Company  

Bejoy spent his career in enterprise technology, including a stretch at IBM working within the Watson group during the years IBM was pushing to lead in AI. When he left IBM, he set out to build AI services on the Watson platform for whichever industry needed them, with no fixed niche in mind.

That changed over lunch. Bejoy, based in Dallas, met Troy Broussard, CEO of the Dallas Housing Authority, to pitch his new company. The meeting was booked for an hour. It ran four. Broussard introduced Bejoy to the public housing world, and after some early research, Bejoy and his CTO settled on a first project: optimizing inspections, a problem Broussard's team was already wrestling with. That early win became the opening into an industry Bejoy says still runs largely on outdated software and paper.

What Bob.AI Actually Does  

Bob.AI is built around a specific rule that governs US housing assistance: no family should pay more than 40 percent of its income toward rent and utilities, and every family pays at least 30 percent, so there is still some stake in the arrangement. Verifying that number is documentation heavy. Bob.AI reads pay stubs and Social Security award letters to work out a family's net income, then helps a housing authority determine how much the government covers and how much the family pays.

Bejoy describes three distinct groups that use the platform. The customer is the housing authority, effectively the city. The users are the housing authority's employees, the members of assisted families, and the landlords who rent to voucher holders. On the voucher side, Bob.AI can create a three-way contract between a private landlord, the government, and a family. On the public housing side, where the government owns the building directly, the contract runs only between the housing authority and the resident family.

None of this, Bejoy is careful to say, replaces the case manager. It is meant to take the paperwork off their desk.

The Families Who Get Their Homes Back Fast  

Bejoy shared a few examples from families and landlords who reached out directly. One involved a household that needed to move because their school district changed. Another needed a new unit fast after finding rats in the one they had. Requests like these, he says, used to take weeks or months to process, sometimes longer. On Bob.AI, a family can submit a request, get a voucher issued, and get matched to a new home within hours or days.

Landlords have noticed the same shift from the other side. Bejoy says landlords have told him that government contracts which used to take several months to finalize now get approved within minutes on the platform, which is part of why some landlords are choosing to rent to more voucher holders rather than fewer. A faster contract also means a unit sits empty for less time while it waits on paperwork, which Bejoy frames as a direct addition to the local housing supply.

Housing Authorities Are Understaffed and Buried in Compliance  

Public housing authorities carry two kinds of pressure at once: too few staff and too much required documentation. Bejoy uses Dallas as his example, where he says a single case manager can be responsible for 800 to 900 people receiving housing assistance. For a housing authority's leadership, Bob.AI functions as compliance software, keeping a running, audit-ready record of how a family's income was calculated and how the payment split between the family and the government was reached. That record stays ready for a routine internal review or an audit brought in from outside.

That combination of automation and documentation is part of why the platform has grown. Bob.AI now works with 24 housing authorities, meaning 24 cities, after starting in Dallas several years ago. Bejoy is careful to note that Bob.AI does not eliminate the need for case managers. It is meant to give them their time back.

"Every Technology Comes to Us Last"  

One story from Bejoy's early days at the Dallas Housing Authority stuck with him. David Sapsody, then the authority's Chief Operating Officer and now retired, sensed Bejoy was still deciding whether to commit to the industry. He told Bejoy that public housing authorities are typically the last to receive new technology: computers, the internet, and cloud tools all arrived there last.

He wanted this time to be different.

Bejoy says he took that conversation to heart. He also describes visiting the homes of families going through inspections, asking directly how the software could help, and coming back with new ideas or, sometimes, confirmation that something is working.

Can Rental Assistance Become a Bridge to homeownership?  

Bejoy's longer term goal for Bob.AI is to help families move from rental assistance into the middle class, with homeownership as the concrete marker of that shift. He points to a structural problem standing in the way: a family's income generally has to rise for a mortgage to become realistic, but rising income is also what phases a family out of rental assistance. Both Bejoy and Dr. Nall agree that closing that gap will likely take more than software alone, including support from employers, financial education, and access to childcare.

By Bejoy's own estimate, only about 0.01 percent of families in the program currently move successfully into homeownership, a figure he offered as a rough sense of scale rather than a precise calculation. He argues that Bob.AI is already sitting on the underlying data, granular income and asset information built up over each family's time in the program, that a homeownership pathway would need, and calls it a solvable problem rather than an easy one.

Staying Inside Strict Federal Guardrails  

Fair housing law already sets firm boundaries against discrimination in rental assistance, and Bejoy says Bob.AI's contracts require compliance with those federal regulations. The newer question, he acknowledges, is whether the AI itself could introduce bias that a fixed regulation was not written to catch. His team monitors conversations across several dimensions, including tone and the decisions the AI makes, to watch for that kind of drift before it becomes a pattern.

For public housing authorities looking to modernize compliance and give case managers their time back, and for families and landlords navigating the voucher process, Bob.AI is live at bob.ai. Visiting the site opens a chat with Bob, where Bejoy says any housing question can be asked directly.

For more conversations with the founders applying AI to problems that affect real communities, subscribe to Lead with AI on your favorite podcast platform.

Quick Answers  

What is Bob.AI? Bob.AI is a web and mobile platform that helps low income families find housing and helps public housing authorities process the paperwork behind that assistance, including income verification, voucher approval, and landlord contracts.

Does Bob.AI replace case managers? No. Bejoy Narayana says the platform is built to take on documentation and compliance work so case managers can spend more of their time directly with the families they serve.

How does Bob.AI calculate what a family pays? The platform is built around a rule where no family pays more than 40 percent of its income toward rent and utilities, and every family pays at least 30 percent. Bob.AI reads documents like pay stubs and Social Security award letters to help calculate a family's net income toward that split.

Who pays for Bob.AI? Public housing authorities pay for the platform. Landlords and families do not pay to use it.

Who is Bejoy Narayana? Bejoy Narayana is the Founder, CEO, and Product Architect of Bob.AI. He previously worked in enterprise technology, including time at IBM within the Watson group, before a meeting with the CEO of the Dallas Housing Authority led him to build an AI platform for public housing.

Follow or Subscribe to Lead with AI Podcast on your favorite platforms: Website: LeadwithAIPodcast.com | Apple Podcasts: Lead with AI | Spotify: Lead with AI | YouTube: @LeadwithAIPodcast | Facebook: Lead with AI | Instagram: @LeadwithAIpodcast | TikTok: @LeadwithAIpodcast | Twitter (X): @LeadwithAI

Follow Dr. Tamara Nall: LinkedIn: @TamaraNall | Website: TamaraNall.com | Email: Tamara@LeadwithAIPodcast.com

Follow Bejoy Narayana (Founder, CEO & Product Architect, Bob.AI): LinkedIn: @bejoynarayana | Website: bob.ai