Software engineer across full stack, ML, and the systems underneath. I'm at my best turning messy, ambitious problems into software that's fast, reliable, and genuinely usable.
Open to full-time Software, ML, and AI Engineering roles across the US. Let's build something together.
01 — open source
02 — projects
The index that beats the libraries
Everyone imports a vector search library. I wrote my own in C++ and made it win. It answers nearest neighbor queries 1.8 times faster than hnswlib and 2.5 times faster than FAISS, at 0.999 recall, on a single thread.
view case study →
The LLM gateway that remembers
A gateway that remembers what your model already said. It matches prompts by meaning, not exact text, so 78 percent of requests skip the model entirely and the typical answer comes back in 44 milliseconds. The memory runs on Proxima.
view case study →
A policy gateway between AI agents and their tools
A Rust gateway for the Model Context Protocol. It sits between an agent and its MCP servers, refuses any tool call its policy did not permit, reads tool output before the model does to redact credentials and fence off injected instructions, and writes an audit record for every decision. Default deny, 224 tests, CI on Linux, Windows and macOS.
view case study →
A storage engine that never loses an acknowledged write
A multithreaded block storage engine in C++17 for Linux: block reads and writes, atomic batches, snapshots and CRC-32C checksums, with crash recovery from a write-ahead log and double-buffered checkpoints. Validated by 35 fault-injection scenarios that crash it at every file-system call, flip every byte of its metadata, and kill it mid-write with SIGKILL.
view case study →
Media over QUIC, WebRTC and HLS on the same bad network
A single-command testbed that runs three live media delivery models side by side under controlled network impairment: a five-region Media over QUIC relay tree, a WebRTC SFU and HLS. MoQ reached first media fastest in seven of eight network profiles, and a controlled A/B showed that pinning the publisher's bitrate cut stalls by 72 to 83 percent with goodput unchanged.
view case study →
The TA that never sleeps
The grading platform I designed after 400+ Java and Python submissions landed on my desk as a TA. Six specialized agents grade in parallel under Temporal orchestration, a confidence gate routes only the shaky grades to a human, and it runs local-first, so student work never leaves the machine.
view case study →
A parallel compiler that shrinks circuits
A C++ compiler that turns SystemVerilog into a smaller gate netlist and proves it still computes the same thing. Six optimization passes cut an 836K-gate benchmark by 80 percent, and optimizing independent logic regions on a work-stealing pool runs 3.0 to 3.5 times faster on 16 threads, with identical output at any thread count.
view case study →
Catching driving regressions before a build ships
A validation platform for autonomous-driving software builds. It replays the same seeded scenarios through a known-good build and a release candidate, compares what the two did, classifies every regression, and turns each cluster of failures into a report an engineer can reproduce with one command. One 100,000-drive campaign runs end to end in 52 seconds.
view case study →
A build farm that never compiles the same thing twice
A distributed build engine for C and C++. A gRPC coordinator schedules the build graph out of PostgreSQL, workers pull tasks and survive crashes, and every result is cached by content hash. Over a scripted 20-step edit workload it ran the compiler 84 percent fewer times than make, and never recompiled code it had already built.
view case study →
Synthesizable Verilog from a structured description
A C++ generator that turns JSON design descriptions into synthesizable Verilog: module hierarchy, ports, connectivity, inserted pipeline and adapter logic. Graph checks run before anything is written. 1,200 generated configurations all synthesize in Yosys, and on 1,000 deliberately broken descriptions invalid output drops from 766 files to none.
view case study →
Clock, reset and CDC checks with targeted testbenches
A static checker for clock domains, reset trees, synchronizers, clock gating and cross-domain paths in Verilog, with 16 rules and a fix for each finding. For every violation it writes and runs a targeted testbench and draws the waveform at the first hazard. It found all 570 bugs injected into 870 generated designs with no false alarms, and the tests reproduced 89 percent of them.
view case study →
Malware Mugshots
My thesis. I turned malware files into pictures and taught three vision models to recognize the family on sight. They vote together and agree 94 percent of the time across 17 families. It became a first-author paper, now on arXiv.
view case study →
Describe a vibe, get movies
A two-stage hybrid movie recommender, built the way production systems are. A two-tower neural model retrieves candidates through Proxima, my own C++ vector search engine; a neural ranker reorders them. It answers free-text requests like 'a slow-burn sci-fi like Arrival but funnier' and includes movies in theaters this week.
view case study →
Watch the search walk
Vector search does not scan your data, it walks a graph. This runs the real HNSW algorithm on points you can see and replays the walk hop by hop, so you can watch a query fall through the layer stack and land on its neighbors. Click to move the query, drag the sliders, break it on purpose.
view case study →
Let the model learn a task once, then replay it without one
Computer-use automation for back-office web apps that have no API. A model works out how to finish a task in a legacy app once, and the run is recorded as a typed capability that replays deterministically with no model in the loop, tells a business outcome apart from a broken app, and hands the live browser to a person when it meets something it does not know.
view case study →
Live mini games for two people far apart
A real-time multiplayer PWA I built for my long-distance girlfriend and me. Surprise games drop onto both phones at random times when we're both awake, we play live for a minute, and a separate AI agent writes a new game every day. One Durable Object per room, Web Push written from scratch, and it runs for $0 on Cloudflare's free plan.
view case study →
Six Degrees
A social platform that thinks in relationships. It keeps 50 thousand connections in a graph database and answers who knows who in two hops, with profiles and content living in their own store alongside.
view case study →
Cited, Not Guessed
A study assistant that refuses to make things up. Every answer is pulled from real source material and arrives with citations, and it stays quiet when it genuinely does not know.
view case study →
What makes a song a hit
A million messy music records that shared no common keys. I joined them with C++ MapReduce on a Hadoop cluster, then asked the clean data what actually makes a song popular.
view case study →
Telling a fall from sitting down
Fall detection and activity recognition from raw smartphone accelerometer and gyroscope data, comparing classical models, deep networks on raw windows and transfer learning from spectrograms. A 1D CNN detected falls with 99.35 percent accuracy, and held 98.1 percent when every test subject was unseen.
view case study →
Show of Hands
Contactless identity from the palm of your hand. Four models study every scan and vote on the verdict, and together they reach 99.75 percent AUC.
view case study →
Who you are, from how you walk
Identifying and verifying people from the way they walk, using a phone's accelerometer and gyroscope. A feature-based network identified walkers with 90 percent accuracy and verified them at a 3 percent equal error rate; an LSTM on the raw signal verified at 2 percent.
view case study →
Zero to shipped in a week
An AI patient assistant that talks back, in text and in voice. I took it from an empty folder to a shipped product in under a week, working entirely AI native.
view case study →
TCP, from Scratch
I rebuilt TCP by hand over raw sockets. The handshake, the sliding window, the retransmissions when packets vanish, all of it, with a live dashboard watching every packet move.
view case study →
03 — publications
arXiv · cs.CR · 2609.26281
S. Lokhande, F. Di Troia, M. Jurecek, M. Stamp
Appearing as a chapter in Artificial Intelligence for Cyber Defense in Emerging Threats (Springer, 2027)
Malware binaries rendered as images across eight conversion strategies, then read by three complementary feature tracks: handcrafted HOG and Haralick descriptors, frozen VGG16, ResNet50 and ViT-B/16 embeddings, and a custom CNN. Fifteen selected voters beat every individual model.
04 — journey
BE, Computer Engineering
Aug 2019 — May 2023
MS, Computer Science
Aug 2024 — May 2026
Software Engineer Intern
Nov 2021 — Jan 2022
Software Engineering Intern
Jun 2022 — Oct 2022
Graduate Assistant
Aug 2025 — May 2026
05 — research
Conference
Comprehensive Big Data Analytics Dashboard on Chats
Presented at ICICN 2023, Mumbai, India
Scholarship
Spartan Scholar
San José State University · awarded for academic excellence
Felicitation
Consistent High Rank
University of Mumbai · felicitated for holding a top rank across 7 straight semesters
06 — skills
07 — about
I'm fascinated by the gap between a great idea and a system that actually holds up at scale. That gap is where I like to work, building software where performance, reliability, and intelligence meet.
I recently finished my Master's in Computer Science at San José State University. My research on image-based malware classification became a first-author paper, now on arXiv and appearing as a Springer book chapter, and somewhere in that work I fell for building intelligent systems that survive the real world, not just the lab.
Before grad school I was a software engineering intern on an EdTech platform serving 70K+ users, building Python microservices, chasing down latency, and shipping features that thousands of learners actually used. The problems I enjoy most sit right where tough engineering constraints meet a simple, intuitive experience.
Outside of work I like taking systems apart to really understand them. Lately that meant writing a vector search engine from scratch in C++ and an LLM gateway that runs on top of it, just to learn how modern AI infrastructure works under the hood. I ship something small almost every day; building is how I learn.
I'm looking for full-time Software, Machine Learning, and AI Engineering roles where I can build scalable systems that solve real problems. Based in the San Francisco Bay Area and open to relocation.
