[ Graduate Coursework & Applied Labs ]

Coursework & Systems Labs

Graduate coursework and systems engineering labs across dynamic programming, networks, distributed consensus, relational schema design, and SQL query optimization at the University of Georgia.

Subject 012 Projects

Algorithms

Dynamic programming and statistical models applied to biological sequences and optimization problems.

Focus Areas:Hidden Markov models, Viterbi log-space decoding, and RNA secondary structure folding.
Graduate Course Project
Demo

Gene Finding with Hidden Markov Model

A two-state hidden Markov model that labels every base of a DNA sequence as background or gene, decoded with the Viterbi algorithm.

My Contribution: I implemented the statistical model, sequence generator, and Viterbi decoder in log space.
PythonPythonNumPyDynamic ProgrammingHMM
Implementation Highlights
  • Two hidden states: uniform background vs. GC/AT-skewed gene emissions.
  • Viterbi dynamic programming running in log-space to prevent numerical underflow on long sequences.
  • Synthetic gene planter generator used to benchmark and verify decoding accuracy.
Algorithm & System Mechanics
  • Maintains the optimal log-probability of ending in each state for every sequence position.
  • Back-pointer table tracks maximum likelihood transitions, walked backwards to recover the state sequence.
Architecture verification: Runs client-side statistical models on synthetic genomic sequences.
Graduate Course Project
Demo

RNA Secondary Structure Prediction

A dynamic-programming predictor that pairs RNA bases to maximise base-pair stability score, outputting structures in dot-bracket notation.

My Contribution: I implemented the scoring matrix, interval table fill, and recursive traceback.
PythonPythonDynamic ProgrammingBioinformatics
Implementation Highlights
  • Base pair scoring: A-U (2), C-G (3), and G-U wobble pair (1) with minimum loop constraint of 4 bases.
  • Optimal substructure: for each substring, evaluate unpaired bases, end-pairs, or bifurcation splits.
  • Full traceback reconstructing valid secondary structure dot-bracket strings.
Algorithm & System Mechanics
  • Fills the dynamic programming table by increasing substring length so all subproblems are ready.
  • Top-right cell holds the maximum global pairing score; traceback resolves the paired positions.
Architecture verification: Score-maximizing pairing algorithm based on classic Nussinov-style dynamic programming.
Subject 023 Projects

Computer networks

Low-level network engineering: packet crafting, custom protocol servers, TLS range downloads, and latency analysis.

Focus Areas:UDP sockets, DNS wire-format parsing, DoH transport, OpenSSL multithreading, and network measurement.
Course Project
Demo

DNS Forwarder with Blocking & DoH

A DNS forwarder that intercepts queries, answers NXDOMAIN for blocklisted domains, and forwards legitimate requests via UDP or DNS-over-HTTPS.

My Contribution: I wrote the forwarder in Python, and a companion C variant that parses DNS wire-format by hand.
PythonPythonUDP SocketsScapyDNS-over-HTTPS
Implementation Highlights
  • Listens on UDP port 53, dissecting 12-byte DNS headers and length-prefixed domain labels.
  • Synthesizes immediate NXDOMAIN packets for deny-list matches without touching upstream resolvers.
  • RFC 8484 compliant base64url DNS-over-HTTPS upstream relay support.
  • Real-time access audit logging with ALLOW/DENY decision records.
Algorithm & System Mechanics
  • Inspects questions in the wire packet; forwards transparently or crafts an authoritative denial response.
  • Encapsulates raw wire bytes into HTTPS payloads for encrypted DoH resolution.
Architecture verification: Compliant wire-format parsing with local simulated upstream resolver.
Course Project
Demo

Multi-threaded HTTPS Parallel Downloader

A robust C utility that queries file metadata via HEAD requests, splits files into parallel byte ranges, and reassembles segments over TLS.

My Contribution: I built the multithreaded architecture, range request dispatcher, and segment stitcher in C.
CCPOSIX ThreadsOpenSSLHTTP Range
Implementation Highlights
  • Queries remote Content-Length using TLS HEAD requests, calculating chunk boundaries.
  • Spawns POSIX worker threads per byte range with independent OpenSSL sessions.
  • Per-part exponential retry handling (MAX_RETRIES = 3) with isolated disk writes.
  • Atomic sequential file reassembly upon thread join completion.
Algorithm & System Mechanics
  • Leverages HTTP Range headers ('bytes=start-end') to allow non-contiguous concurrent downloading.
  • Independent segment files eliminate thread lock contention during high-speed transfers.
Architecture verification: POSIX pthread and OpenSSL architecture verified against standard HTTP range servers.
Course Project
Demo

Ping & Traceroute Diagnostic Suite

Automated network measurement tooling that captures latency distributions, parses multi-hop traceroutes, and computes per-hop statistical variance.

My Contribution: I wrote both measurement tools, data parsing regexes, and statistical summarization.
PythonPythonICMPMatplotlibStatistics
Implementation Highlights
  • Automated ping suite calculating mean, median, min, max, and jitter distributions.
  • Multi-iteration traceroute parser grouping hop-by-hop latency samples to expose routing instability.
  • Box-plot outlier detection revealing congested intermediate autonomous systems.
Algorithm & System Mechanics
  • Extracts per-probe timestamps, groups latency metrics across iterations, and exports structured JSON.
Architecture verification: Parsed from real traceroute datasets collected across commercial network hops.
Subject 034 Projects

Distributed Computing System

Core distributed systems paradigms: concurrency control, message-passing protocols, persistent multicast, and hash rings.

Focus Areas:TCP concurrency, ReadWriteLock synchronization, persistent stateful coordinators, and consistent hashing.
Pair Project with Tejas D. ShindeJan – Feb 2025
Demo

Multi-threaded FTP Client & Server

A Java client and threaded server implementing a custom FTP protocol over TCP sockets with stateful working directory management.

My Contribution: We co-authored the threaded connection handler, stream protocol parser, and virtual file operations.
JavaJavaTCP SocketsMultithreadingI/O
Implementation Highlights
  • Concurrent client support via thread-per-connection pool with isolated session states.
  • Framed length-prefixed stream transfer protocol for unambiguous EOF detection.
  • Full command set: GET, PUT, LS, CD, MKDIR, PWD, and graceful QUIT.
Algorithm & System Mechanics
  • Sends byte-count headers prior to raw payload transfer on streaming TCP sockets.
Architecture verification: Faithful recreation of TCP protocol mechanics and socket session handling.
Pair Project with Tejas D. ShindeFeb 2025
Demo

Concurrent FTP with Granular File Locks

An advanced FTP server featuring non-blocking background transfers, per-file ReadWriteLock concurrency, and an out-of-band terminate channel.

My Contribution: I architected the per-file locking system, async transfer worker pool, and dual-port terminate interrupt.
JavaJavaReadWriteLockConcurrentMapSockets
Implementation Highlights
  • Dual-port architecture: Port A handles data commands, Port B receives out-of-band terminate requests.
  • Granular per-file ReadWriteLock: multiple concurrent readers permitted, writers get exclusive access.
  • Asynchronous background task tracking with unique transfer task identifiers.
  • Periodic chunk-level cancellation checks with automatic partial-file rollback.
Algorithm & System Mechanics
  • Maintains lock tokens in a ConcurrentHashMap keyed by canonical file path.
  • Secondary control port allows instant kill commands without head-of-line blocking.
Architecture verification: Recreates Java concurrent locking semantics and dual-socket command routing.
Individual ProjectMar 2025
Demo

Persistent Multicast with Time-Window Coordinator

A distributed group-communication protocol where a centralized coordinator buffers messages for offline members within a configurable time window.

My Contribution: I designed and implemented the coordinator state engine, participant protocol, and message time-to-live queues.
JavaJavaTCP SocketsDistributed SystemsState Machines
Implementation Highlights
  • Dynamic group membership: register, deregister, temporary disconnect, and reconnect commands.
  • Time-bounded message persistence (threshold td) ensuring offline participants catch up upon reconnection.
  • Recipient snapshots captured at broadcast time to prevent retroactive leakage to late joiners.
  • Audit event logging on every individual participant.
Algorithm & System Mechanics
  • Coordinator records message recipients at the moment of send; buffers unexpired messages in memory.
  • Reconnecting nodes receive buffered unexpired messages addressed to them.
Architecture verification: Simulates time-window buffered multicast semantics and connection lifecycle events.
Individual ProjectMar – Apr 2025
Demo

Consistent Hashing Distributed Key-Value Ring

A decentralized key-value storage service partitioned across a 1024-position consistent hash ring with dynamic node joining, graceful departures, and key handovers.

My Contribution: I developed the bootstrap coordinator, ring routing logic, and dynamic key range migration.
JavaJavaConsistent HashingTCP SocketsRing Topology
Implementation Highlights
  • 1024-position logical ring where keys map to the nearest clockwise active node.
  • Dynamic membership: new nodes claim key range (predecessor, id] with atomic handover.
  • Graceful node departure: exiting node forwards stored pairs to its successor before sign-off.
  • Request routing: query routing through bootstrap index with path tracing.
Algorithm & System Mechanics
  • Ring partitioning confines membership churn to adjacent neighbor nodes.
  • Key migrations occur only between the entering/exiting node and its immediate successor.
Architecture verification: Full consistent hashing ring algorithm verified across hundreds of join/leave cycles.
Subject 042 Projects

Database Management System

Relational database schema modeling, role-based application workflows, and analytical SQL query engineering on PostgreSQL.

Focus Areas:PostgreSQL 3NF schemas, role-based access control, CTEs, complex joins, and query plan optimization.
Team of Three: Harmanpreet Singh, Dipendra Shrestha, Tejas D. ShindeApr 2025
Demo

Enterprise Multi-Role Property Management

A comprehensive full-stack property management platform featuring four role-specific dashboards, lease agreement tracking, maintenance ticketing, and payment logs.

My Contribution: I architected the frontend application, authentication flows, role-based access control (RBAC), and interactive dashboards.
PostgreSQLPostgreSQLDrizzle ORMDrizzle ORMTypeScriptTypeScriptExpressExpressNuxt 3Nuxt 3Vue 3Vue 3
Implementation Highlights
  • Normalized PostgreSQL schema modeling tenants, owners, properties, leases, payments, and work orders.
  • Role-based permission gating for Administrators, Property Managers, Property Owners, and Tenants.
  • End-to-end maintenance ticketing lifecycle: Pending -> In-Progress -> Resolved.
  • JWT session management and Drizzle ORM query optimizations.
Algorithm & System Mechanics
  • Role context governs accessible data queries and mutation actions.
  • Relational foreign keys bind tenant rental contracts directly to maintenance billing ledgers.
Architecture verification: Production-grade database schema and multi-role UI recreation.
Individual Practice & CurriculumFeb 2026
Demo

38 SQL Exercises on PostgreSQL Analytics

A curriculum of 38 progressive SQL queries written for PostgreSQL over a normalized job-postings schema covering aggregation, windowing, subqueries, and CTEs.

My Contribution: I authored all thirty-eight analytical queries, edge-case fixes, and query plan notes.
PostgreSQLPostgreSQLSQLRelational AlgebraCTEsAnalytics
Implementation Highlights
  • Normalized star/snowflake schema: job_postings_fact, company_dim, skills_dim, and skills_job_dim.
  • Complex aggregations: salary trends by skill combinations, remote posting ratios, and year-over-year demand.
  • Advanced techniques: recursive Common Table Expressions (CTEs), CASE branching, and date-truncation grouping.
  • Comparative query analysis highlighting index utilization and NULL handling nuances.
Algorithm & System Mechanics
  • Bridges fact and dimension tables through associative link tables to model many-to-many skill requirements.
  • CTEs modularize multi-pass aggregation for clear query plan evaluation.
Architecture verification: PostgreSQL compliant queries authored and benchmarked against relational datasets.

Interested in full-scale applications and research?

Explore live production platforms, published research papers, and developer tooling.