TECHNICAL PRODUCT MANAGER
I FIX
AMBIGUITY.
I work at the intersection of product thinking, execution, and system risk. I ensure every build delivers value and thrives in user reality.
I started in the trenches.
I built my foundation in engineering. I mastered the subtle ways systems scale, break, and evolve under pressure.
This technical DNA dictates how I lead. I don't just write requirements; I visualize the logic flow and anticipate edge cases before they become expensive problems.
CLARITY IS
PREDICTABILITY.
I begin with systems that won't break.
I prioritize the core experience. A feature that works 100% of the time is worth more than ten that fail predictably.
Changing direction is cheap now.
I surface assumptions early. I find the friction in the logic before we write a single line of production code.
[ EXECUTION_CAPABILITIES ]
01. DECISION CLARITY
I turn vague "user needs" into specific moments of struggle, removing weeks of wasted work.
02. RELEASE CONFIDENCE
Quality is a design input. I define acceptance criteria that actually mean something.
03. EXPLICIT TRADE-OFFS
I put speed, stability, and cost on the table—and force a choice before conflict starts.
MY_JOURNEY
HOW I LEARNED TO BUILD THINGS THAT WORK.
I engineered my path to product management by mastering every discipline that defines the craft.
Started Where Most PMs Rarely Go
As a technical writer at TechTunes, I broke down complex technology for thousands of readers. I learned that true product mastery means taking the impenetrable and making it clear. Clarity became my strength. When you can explain a system simply, you understand it well enough to ship it.
Inside the CodeWent to the Source of All Product Truth
At Microsoft, I shifted from explaining tech to understanding the people who use it. As a Customer Insights Intern, I learned to speak the language that bridges engineering and empathy: data. I discovered that great products live in the space between what developers create and what users actually need.
User RealityReal-World Execution Under Pressure
In pharmaceutical and retail businesses, I transformed manual, bottlenecked operations into streamlined digital systems. I built POS platforms and marketplaces from scratch. I learned that product management is really about making the right call when resources are tight, timelines are brutal, and failure is costly.
Building from ZeroProved I Could Master Any Complex System
I deliberately tackled the Bangladesh Civil Service exams. I treated it like reverse-engineering a legacy codebase. I found the pattern, decoded the architecture, and executed. I won. This was proof of principle: any complex system can be mastered through methodology and strategic focus.
Pattern RecognitionWhere Strategy, Quality, and Execution Converge
At ELO, I started as a Quality Assurance Lead, rebuilding the definition of reliability. I architected automation pipelines and release processes that achieved near-perfect uptime. Then I moved to Technical Product Manager, where I now lead strategy across a massive portfolio. I merge the empathy of my early years, the technical rigor of my QA days, and the strategic thinking from my entire journey.
Full-Stack PMI am a PM who deliberately mastered every discipline, then synthesized them into a singular skillset. I build systems that work: reliably, strategically, and at scale.
[ RECENT_LOG ]
TURNING CHAOS INTO CERTAINTY.
AI Sound Matching Desktop App
Critical Revenue RecoveryTHE CONTEXT
An AI sound matching platform had built their core technology, but faced a critical challenge: their Desktop Agent App was their primary revenue channel. Music producers worldwide relied on it to integrate AI-powered sound matching directly into their production workflow. When the app worked, it saved producers hours of manual searching. When it crashed, they lost trust and cancelled subscriptions.
THE PROBLEM
The app was crashing frequently during audio processing, especially when producers analyzed large music libraries. Support tickets were flooding in. Churn was rising. The engineering team had optimized for feature parity with the web version, but the desktop app architecture was fundamentally broken. It loaded entire audio libraries into memory at once, causing memory leaks that froze the application. Producers with 10,000+ track libraries could not use the product at all. The business was at risk of losing their most valuable customers.
MY SOLUTION
I froze all new feature development and assembled the team for a focused stability sprint. I ran performance profiling sessions using real customer music libraries to identify exactly where memory was leaking. The data showed the audio analysis pipeline was the bottleneck. I worked with engineering to redesign the entire processing architecture from batch loading to streaming. We implemented chunk-based processing with automatic memory cleanup after each batch. Then I introduced a stability gate: no release could ship unless it passed 24-hour stress tests processing 10,000+ track libraries without crashes.
THE IMPACT
92% crash reduction within two weeks. Large library processing became 5X faster. Most importantly, we stabilized the critical revenue channel. Customer renewal rates recovered, and support tickets related to crashes virtually disappeared. The product became reliable enough to power the core business.
European Headhunting Platform
AI Trust & AdoptionTHE CONTEXT
A European headhunting platform had invested heavily in AI-powered candidate matching technology to connect companies with executive talent faster. The algorithm worked technically and matched candidates to job requirements with high accuracy. Yet recruiters were still manually reviewing every single candidate before presenting profiles to clients. The entire process took weeks, defeating the purpose of automation.
THE PROBLEM
Recruiters did not trust the AI. The system gave candidates percentage match scores, but recruiters could not understand why a candidate scored 87% versus 92%. Executive hiring is not about keyword matching. It involves nuanced judgment about career trajectory, cultural fit, leadership style, and relationship potential. The AI treated hiring like a search engine problem when it was actually a trust and context problem. Recruiters bypassed the expensive technology entirely and went back to Excel spreadsheets.
MY SOLUTION
I spent time with 15 senior recruiters to understand how they actually evaluated candidates in practice. I discovered they weighted soft signals like career progression patterns and cultural alignment far more than resume keywords. I worked with the team to completely redesign how the matching model presented information. Instead of showing scores, we surfaced context: why this candidate matched, what made their background relevant, what potential concerns existed. We built recruiter-facing summaries that explained the match in human terms. Then we added feedback loops so the model could learn from real placement outcomes, not just theoretical matches.
THE IMPACT
68% faster shortlisting process. Recruiters started using the platform actively, with adoption increasing 4X. Most importantly, successful placements improved by 35% because the AI was now helping recruiters make better decisions instead of being ignored. The expensive technology investment finally delivered business value.
Healthcare Scheduling Platform
System Complexity & ScaleTHE CONTEXT
A healthcare appointment scheduling platform served clinics and patients with a reliable booking system. It worked perfectly for single-doctor practices. Then multi-specialty clinics started onboarding, and everything broke. Double-bookings became common. Patients showed up for appointments only to find doctors unavailable or equipment already in use. Clinics were receiving complaints. Trust in the system was eroding fast.
THE PROBLEM
The scheduling logic made a critical assumption: one doctor equals one time slot. This worked fine for simple practices. But multi-specialty clinics share resources like diagnostic equipment, examination rooms, nurses, and support staff. A cardiologist might need an ECG machine, but so might the pulmonologist. The system allowed both doctors to book the same equipment at the same time because it only checked doctor availability. The quality assurance process tested single-doctor scenarios perfectly but missed the complex resource conflicts that broke real clinics.
MY SOLUTION
I mapped the complete resource dependency tree for multi-specialty clinics by interviewing clinic administrators. I identified six different types of resource conflicts the system was blind to: equipment, rooms, support staff, pre-procedure prep time, post-procedure recovery space, and doctor handoff requirements. I worked with engineering to implement resource-aware scheduling with conflict detection at booking time, not after confirmation. Then I built an automated test suite generator that created multi-resource scenarios dynamically, ensuring we could never regress on this complexity again.
THE IMPACT
Zero double-bookings after launch. Multi-specialty clinic adoption increased by 250% because the product finally worked for complex practices. We achieved 95% test coverage for edge cases that previously caused failures. The platform became reliable enough to handle the healthcare market segment that mattered most for growth.
WHY I WORK WITH ENGINEERS?
I lead through clarity. I respect constraints. I bring clear framing and reduced back-and-forth, so good engineers can move at full speed.
Fewer rollbacks
Optimized
Fewer urgent hotfixes
Reduced
Low support escalations
Controlled
High signal from QA
Verified
