Search current work, talks, writing, public links, and local tools.
Builder, operator, customer loop
[SURFACE 01]
Texts, tickets, reviews, calls, renewals
Arjun Kannan
I build software for teams that keep hearing the same customer problems and need a better way to act on them.
At ResiDesk, that means resident texts, reviews, calls, renewals, and support threads become answers, owner context, and work a property team can actually pick up.
Talk to your customer is still business 101. The harder version is when the customer has already talked to you 10,000 times and the pattern is buried.
ResiDesk is where I am spending most of my time now. Before that it was outcome-based lending at Climb Credit and advisor tools at BlackRock. The throughline is pretty simple: understand the customer, then build the thing that helps a person make the next call.
I care less about the demo and more about the next morning: the queue is messy, the edge cases are real, and nobody is grading the product except the person who has to use it.
Current work: ResiDesk. Before that: Climb Credit and BlackRock. I try to make the important thing easier to see before the team gets buried in one-off work.
Most teams are not missing feedback. The feedback just loses its owner before it changes rent, renewals, maintenance, staffing, or the product.
The next morning matters
I want to know what someone still uses when the meeting is over, the queue is messy, and a customer is waiting for a real answer.
Do not make the team guess
A good system makes the team faster without making them reckless. It shows the policy, history, tone, uncertainty, and who should take the next step.
This is why ResiDesk makes sense to me
Climb, BlackRock, the talks, and the writing all point to the same thing: useful software gives the person doing the work a better read on the customer.
Start here
[PATHS 03]
Start with what you need
Most people come here for one of three reasons. Start with the one closest to your problem.
We help property teams answer residents, see what is happening across the building, and get the right issue to the right person without rereading the whole history.
Looking for
Teams that do not want another AI support chat.
The best conversations start with operators who already care about retention, NOI, workload, maintenance, and resident trust. The problem is volume, not caring.
Writing about
What happens after the answer.
AI made it cheap to produce a plausible reply. I still want to know: who owns the next step, what do they know, and does the resident have to explain it again?
Not useful
Replies that go nowhere.
If the reply goes out and nothing changes, I do not trust the product yet. That is not a finished job. It is a nicer inbox.
Work before ResiDesk
[WORK 04]
A few things I have worked on
CompanyWhat I worked onWhat happenedMore context
Climb CreditI was CTO and CPO, trying to put outcomes into product, data, and underwriting.Annual loan volume grew from $1M to $300M while the product got closer to the student's real outcome.TechCrunch
BlackRockI worked on product and engineering where interface quality mattered because real money was behind the decision.The advisor analytics product reached $40M ARR in year one.Work history
ResiDeskI am co-founder and product lead, focused on making resident messages useful before they turn into churn, workload, or NOI pain.Law360 covered a reported 7% lift tied to acting on resident feedback earlier.Law360
How I think
[MODULE 04]
How I work
My default is simple: talk to the customer, make the work visible, remove load from the team, and see whether anyone still uses the thing when the room gets busy.
01 / Customer
Talk to the customer first.
If you have customers, understanding them is business 101. In housing, the hard part is hearing enough residents without burying one person in every text, review, and ticket.
02 / Context
Show me the actual work.
I do not learn much from abstractions by themselves. Give me the stakes, the strange edge cases, and the person who has to live with the outcome.
03 / Adoption
Demos are not adoption.
I learned this early at BlackRock: a prototype can win the room and still lose to the spreadsheet the next morning. The test is what people reach for when the meeting is over.
04 / AI
Shorten the distance.
I do not need AI to do everything. I need it to move something from stuck to almost done while the person still owns the judgment.
05 / Candor
Be direct about the system.
The best teams can say what is broken without turning it into blame. Believe the pain first, then find what actually caused it and who can change it.
06 / Team
Hire people who can work through mess.
The best people I have worked with can walk into a messy situation, find the few facts that matter, and move without waiting for a perfect script.
AI deployment
[MODULE 05]
What happens after the demo?
AI made it easier to sound useful. It did not make it easy to change work. I care about the policy, risk, handoff, trust, and queue that show up after the sales call.
Deployment note
The day after the demo is where the product starts.
I am most useful when a team has a real deployment problem: the demo worked, but now the customer's actual day is pushing back.
01Find the real job
Sit with the people doing the work. Watch what they do when the system is slow, strange, or politically hard.
02Name what changes
Tell me whether it saves time, reduces risk, protects retention, frees capacity, or earns trust. If nothing changes, we are admiring a demo.
03Design around the job
Pick the model late. First decide the context, tools, permissions, evals, review points, and the handoff into the day people already have.
04Test the ugly cases
Run the ugly cases. Show the misses without drama. Measure whether the work moved faster, got safer, or reached the right person sooner.
05Bring field learning back
The best sales material comes from what actually happened: the real objection, the before state, the adoption metric, and the sentence the buyer can repeat honestly.
$40M ARRBlackRock advisor analytics, first year
$300MAnnual loan volume at Climb Credit
100+Startup investments and founder reads
AI product work
[MODULE 06]
After the demo, the real work starts
Capability is not the finish line. The person buying it, using it, reviewing it, and explaining it in a meeting are usually worried about different things.
Field note
Find the work, show the help, and make ownership clear.
I have watched a good demo lose to a spreadsheet. I have also seen hidden information become a product once it reached the person making the decision.
01
Customer-work discovery
Find the narrow place where AI changes the day, not the slide. Talk to the buyer, the user, and the person who gets paged when it breaks.
02
Deployment practice
Make deployments repeatable: scope the job, staff it, test it, review risk, roll it out, and bring field learning back to product.
03
Sales from what actually happened
Help GTM say the true thing: what changed, why the buyer cared, what broke, and what survived the rollout.
04
Executive translation without hand-waving
Make the executive version clear without flattening the technical truth. I care about the number, the risk, the owner, and what happens next.
Resident messages
[MODULE 07]
How one resident message becomes work someone owns
[STEP 01 / LISTEN]
Start with what the resident actually said.
A useful system starts with ordinary, expensive stuff: the broken washer, the pet policy question, the Wi-Fi complaint, the package-room mess, the first sign someone may not renew.
Daily textsReal complaints
ResiDesk
[LOOP 08]
Why housing is worth working on
Before
Resident feedback is scattered.
Texts, reviews, tickets, surveys, renewal notes, and maintenance complaints sit in different places. The owner usually sees the financial result later.
After
The right person sees the issue sooner.
The system keeps enough context to answer, route, report, and show what the building should change before the problem gets expensive.
01Message
Resident text, review, ticket, call, survey, or renewal note.
02Context
Lease, policy, unit, history, tone, and what happened before.
03Owner
The person or team that can actually change the outcome.
04Action
Answer, escalate, repair, explain, or change the policy.
05Report
What owners need to see about retention, NOI, workload, and risk.
About
[MODULE 08]
I did not start in housing
I grew up around research, so the obvious path looked academic. I studied applied physics at Cornell because I liked real systems, messy measurement, and problems where small details could change the answer.
Software came in sideways. I was in an electron microscopy lab and wrote code to make a magnetic-noise setup faster. It saved hours almost immediately. That was the moment software stopped feeling separate from the real work.
The industries changed. The operating question did not. At BlackRock, it meant making institutional tools usable for advisors. At Climb Credit, it meant underwriting against outcomes. At ResiDesk, it means helping housing companies hear residents clearly enough to act.
Physics
Software became the lever.
I wrote a tool in an electron microscopy lab to speed up a magnetic-noise setup. It saved enough time, fast enough, that software started to look like leverage instead of coursework.
BlackRock
Real stakes sharpen the interface.
I came back six months later, re-interviewed, and moved to New York. It taught me that interface quality matters a lot more when real money sits behind the decision.
Climb Credit
Outcomes changed the product.
Instead of asking who looked safest on paper, we asked what happened to earnings after the program. That pushed outcomes into underwriting, product, and data, and annual loan volume went from $1 million to $300 million.
ResiDesk
Housing should know its customer.
Residents tell buildings what is working and what is not every day. The work is making that clear to the owner, useful to the operator, and less annoying for the person living there.
Work history
[MODULE 09]
Where I worked before
The settings changed, but the job stayed similar: understand what the customer is saying inside a messy process, then build the shortest responsible way to act on it.
ResiDesk customer result
7%
Reported lift tied to getting resident feedback into decisions earlier. I keep the source close because this number needs context.
Climb Credit
$1M → $300M
Annual loan volume growth while outcomes became part of the product, data, and underwriting conversation.
Advisor tools
$40M ARR
Advisor-facing analytics product I worked on from zero to $40 million ARR in its first year.
We help rental-property owners and operators understand what residents are asking for across renewals, rent, maintenance, and staffing. The product earns its keep when it changes what happens next: who owns it, what context matters, and whether the work gets done.
We underwrote against a different question: not who looked safest on paper, but what happened to a graduate's earnings. That forced outcomes into the product, data, and underwriting.
The job was turning institutional infrastructure into something advisors could use with clients. Same information underneath, built for the moment when someone had to explain, compare, and decide.
More background
[MODULE 10]
Work, writing, and conversations
Product leadership
I have worked on products where the number had to move.
BlackRock advisor analytics reached $40 million ARR in the first year. At Climb, annual loan volume grew from $1 million to $300 million as outcomes moved into product and data.
If you want more detail, start with the public links.
TechCrunch covered Climb. Law360, HackerNoon, TechTimes, TechBullion, BuiltWorlds, and 20for20 give more context on ResiDesk, applied AI, talks, and operator work.
I write when I have questions. Most pieces come back to the same test: does this help someone finish the work, or did we just make the demo easier to sell?
If a tool does not help someone finish a real task sooner, with less context loss, it is hard for me to care about it.
Understand the job first.
If you do not know what someone is actually trying to do, you are probably just rearranging the screen.
Build the system around the model.
The model is one part. Context, tools, guardrails, evaluation, and the handoff into someone's day decide whether it changes anything.
Demos lie by omission.
What matters is whether people still reach for it mid-work, mid-mess, with no audience and no demo to grade.
FAQ
[MODULE 14]
Fast answers before reaching out
What AI systems do I actually build?
I build AI around work that already exists. At ResiDesk, that means helping property teams answer residents, understand what is happening in the building, and get the right issue to someone who can fix it.
What have I built before ResiDesk?
I worked at Climb Credit and BlackRock. At Climb, I helped annual loan volume grow from $1 million to $300 million. At BlackRock, I worked on a retail analytics product that reached $40 million ARR in its first year.
What do I usually write and speak about?
I usually come back to the same things: agents, evals, product loops, and the gap between a strong demo and something people still use when the day gets messy. Housing makes the point concrete because the customer is already talking.
What is my view on AI?
I care less about whether something looks impressive and more about whether it helps someone make a better call. That usually means getting the context right, testing what good looks like, and keeping a human close enough to stop the system from automating the wrong thing.
Investing
[MODULE 15]
Investing when I can be useful
I have invested in more than 100 startups and mentored through Techstars. I tend to back founders who are close to the problem, close to the customer, and honest about what they do not know yet.
Generic advice is free now. The useful version is specific: here is the customer, here is the constraint, here is the ask, here is the next decision.
You have real customer pain and need it to change product, GTM, or operations.
You are deploying AI where context, evals, handoff, and trust actually matter.
You are a housing operator trying to spot resident issues before they become churn.
Probably not a fit
You want generic AI inspiration without a real customer or job attached.
You need someone to bless a demo that has no owner, metric, or next step.
You want a broad advisory call instead of a specific problem I can help sharpen.
Optional tools
[MODULE 17]
Optional tools for reading the page
01 / Page map
Local visuals ready
See the work as a map.
Pick a view. The graphic runs anywhere. If the browser has a local model, it can add a sharper read on fit or adoption.
02 / Ask this site
Checking browser AI
Ask the page a question.
The answer uses the copy, talks, writing, public links, and tools on this page. Try: "why ResiDesk?", "what is the Tuesday test?", "where should I start?"
Try asking about ResiDesk, founder advice, BlackRock, Climb, writing, or the Tuesday test.
03 / Conversation map
Start with a real question.
04 / Find the relevant bits
Find the sections worth reading.
05 / Transcript lens
Pull the useful parts from a talk.
06 / Useful AI test
Paste an AI idea. See if it has a job.
07 / Tuesday test
Stress-test a demo against Tuesday.
Pick a demo promise and an environment. The simulator shows what has to be true before the thing survives a normal workday.
Run the test to see demo risks, workday pressure, and the adoption check.
08 / Resident messages
Generate a small building readout.
09 / Pattern highlighter
Show the pattern inside the page.
Highlights the words I keep coming back to: customer, context, measurement, follow-through, trust, and demo.
10 / Internal check
Audit the page against its job.
This checks whether the page is clear, useful, and honest about what it is trying to do.
11 / Next pass
Pick the next weak spot.
Pick the part that feels weakest and get a concrete next pass.
