Akshat Lakhera — software engineer building systems where the hard parts matter: latency, correctness, scale, and reasoning.

AKSHAT LAKHERA
Preparing the engine…
AKSHAT LAKHERA
ABOUT

I build systems where the hard parts matter — latency, correctness, scale, reasoning.

Computer Science and Applied Mathematics undergraduate at IIIT Delhi (batch 2023–2027). Two software internships — a defense research lab and a venture-backed startup — plus undergraduate research in parallel and high-performance computing. A Codeforces Specialist with 750+ DSA problems solved, and a finalist in IMC Prosperity 4, IMC Trading's global algorithmic trading competition. The work spans Kafka pipelines, distributed backends, parallel PDE solvers, computer vision, and autonomous LLM agents — because the systems I want to build need all of these to coexist.

EXPERIENCE

Where I've shipped.

Jan – Apr 2026
Software Developer Intern

Shipped a production gamified campus-ambassador platform for 100+ users — dual student and admin portals, live leaderboards, and 20+ REST APIs in TypeScript and Next.js on a 17-table Supabase (PostgreSQL) schema, delivered across Agile sprint cycles. Eliminated concurrency defects under peak load with atomic transactions and row-level locking, enforced secure coding through input validation and parameterized queries, and gated every release behind 433 automated tests across 21 suites.

VIEW →
Aug – Dec 2025
Research Student — Scientific & Financial Simulations

Accelerated the explicit Black-Scholes PDE solver ~10× on multicore CPUs with OpenMP — persistent thread teams, cache-friendly access patterns — and scaled it across nodes with MPI using domain decomposition and ghost-cell exchange, benchmarked against a serial baseline. Implemented and benchmarked Explicit, Implicit, and Crank-Nicolson finite-difference solvers in C++ and Python against analytic prices; exploiting Crank-Nicolson's unconditional stability cut average solve time from 2.69 seconds to ~6 milliseconds at equal-or-better accuracy.

VIEW →
Jun – Jul 2025
Software Developer Intern

Engineered Kafka-based asynchronous communication between Java Spring Boot and C++ microservices over TCP/IP in a distributed, mission-critical defense system — improving processing efficiency by 60%. Trained a YOLOv8 maritime vessel-detection model on HRSC2016 satellite imagery to 0.83 mAP@0.5 via systematic preprocessing and a custom XML-to-YOLO annotation parser, and streamlined C++ Linux service deployments into a CI/CD pipeline, eliminating manual integration errors across test environments.

VIEW →
SELECTED WORK

Projects that taught me something.

TECHNICAL STACK

The tools, and the theory.

Languages
Python C++ C Java JavaScript TypeScript SQL
Technologies
React Next.js Node.js Express Spring Boot FastAPI Kafka REST APIs Microservices PostgreSQL MongoDB Supabase Redis Git CI/CD Linux Vercel RBAC TCP/IP LangChain LangGraph RAG MCP Agentic AI YOLOv8 OpenMP MPI
CS Foundations
Data Structures Algorithms OOP & Design Operating Systems Computer Networks Database Systems Parallel Programming Distributed Systems Multithreading & Concurrency Machine Learning Advanced Programming Open-Source Dev Competitive Programming
Applied Mathematics
Real Analysis Probability & Statistics Stochastic Processes Convex Optimization Statistical Inference Differential Equations Scientific Computing Theory of Computation
RECOGNITION

Receipts.

1458
Codeforces Specialist

Max rating, earned through rated algorithmic contests in competitive programming.

Finalist
IMC Prosperity 4

IMC Trading's global algorithmic trading competition — World Rank 823, All India Rank 175, Algorithmic Rank 278.

100/100
CBSE Mathematics Topper

Perfect score in Class 12 board examination, 2021 cohort.

2
Software Internships

Defense research at DRDO (ISSA Lab) and a venture-backed startup, Grove Growth — shipping production systems at both.

750+
DSA Problems Solved

550+ on LeetCode and 200+ on Codeforces — consistent practice, not a one-time sprint.

SMP
Mentor Recognition Award

Student Mentorship Program, IIIT Delhi (2025–26) — for sustained engagement mentoring first-year students through their academic transition.

WHAT I BRING

Across two internships, I shipped Kafka pipelines that lifted asynchronous processing 60% at a defense lab, and a production platform for 100+ users at a venture-backed startup — every release gated by 433 automated tests. In research, I scaled a Black-Scholes PDE solver ~10× across cores and nodes with OpenMP and MPI, and cut average solve time from 2.69 seconds to ~6 milliseconds. In competition, I'm a Codeforces Specialist with 750+ problems solved, and an IMC Prosperity 4 finalist — All India Rank 175 in a global algorithmic trading field. The CBSE math topper score is from 2021.

I've worked across systems-level C, OpenMP/MPI parallel computing, distributed Kafka, full-stack TypeScript, and autonomous LLM agents — most recently ReturnGuard, a returns-resolution platform where deterministic guardrails, not the model, control the money path. Not breadth for its own sake: the systems I want to build need the hard parts (latency, correctness, scale, reasoning) to coexist. I also mentor first-year students at IIIT Delhi — recognized with the SMP Recognition Award, 2025–26.

Off-keyboard: synthesizer, foosball, logic puzzles, and the financial markets.

GET IN TOUCH

Let's build something.