✦ Software & AI Engineering ✦

Engineering scalable systems &
AI-driven platforms.

Software Engineer specializing in AI-powered platforms, low-latency microservices, and large-scale data pipelines. Proven track record deploying LLM-driven features, vector search, and real-time processing systems using Python, Spring Boot, Kafka, and GCP.

Based in India
Python Spring Boot Big Data LLM Embeddings Vector Databases AI Agents
Samyak Vora
Professional Track

Ontic Technologies

Senior Software Engineer

Oct 2024 — Present · Noida

  • Built a pipeline to extract trending stories from unstructured data by clustering embedding vectors and generating summaries using LLMs, with Guardrails validation streamed through Kafka into Spring Boot.
  • Reduced latency in cartoon image detection by 75% using async inference and batch aggregation in FastAPI with Triton inference server.
  • Implemented similarity search using Qdrant to power natural language-driven universal search across the platform, replacing a legacy Elasticsearch solution.
  • Engineered a vector database-powered module for semantic deduplication and content tagging — reduced duplicate content by ~20% with an average latency of 400ms.
  • Benchmarked and provisioned a Qdrant vector store housing 20M+ points with multi-embedding support for high-throughput semantic workloads.
  • Strengthened platform observability using OpenTelemetry and Uptrace; enforced code quality standards via dependency injection, linters, formatters, and pre-commit hooks.

Infocusp Innovations

Software Engineer

Jul 2022 — Sep 2024 · Ahmedabad

  • Built event-driven pipelines using Apache Beam and Google Pub/Sub to process petabyte-scale logs and operational metrics; modeled data in Cloud Spanner and surfaced insights via BigQuery and Looker Studio.
  • Contributed to FreeRTOS-based firmware for smartwatches on nRF52832, including real-time accelerometer signal processing and custom BLE GATT services for health data synchronisation.
  • Developed CI/CD pipelines to automate testing, linting, and code quality enforcement across engineering teams.

Playpowerlabs

Software Developer Intern

Oct 2021 — Jun 2022 · Gandhinagar

  • Built modular Angular components and implemented Module Federation for improved performance through lazy loading, reducing initial bundle size and improving page load times.
Engineering Builds
01 / HYBRID ARCHITECTURE

Trend Detection Pipeline

Large-scale trend detection system processing ~20M records weekly to extract trending stories from unstructured data. Pipeline spans embedding generation, clustering, LLM summarisation, and Guardrails validation — streamed via Kafka into Spring Boot for real-time insight extraction.

Spring BootKafkagRPCPythonLLMs
02 / INFERENCE OPTIMISATION

Cartoon Detection API

High-performance image classification API optimised for concurrency, built for cartoon image detection. Leveraged asynchronous FastAPI and Triton batch inference to cut infrastructure costs by ~75% and reduce latency by 75% under load.

FastAPIAsyncIOPyTorchTritonBatching
03 / SEARCH & VECTOR

Semantic Deduplication

Vector-based deduplication engine using Qdrant — delivering high-speed similarity search and scalable content filtering at 400ms average latency across a corpus of 20M+ embedded points. Reduced duplicate content by ~20% and powers natural language-driven universal search, replacing legacy Elasticsearch.

QdrantVector SearchSimilarity SearchElasticsearch
04 / FIRMWARE

Health Wearable Firmware

Embedded firmware for health-focused smartwatches on nRF52832 using FreeRTOS. Implemented power-efficient scheduling, real-time accelerometer signal processing, and custom BLE GATT services for live health data synchronisation.

CFreeRTOSnRF52832BLEAccelerometers
Expertise

Backend & Frameworks

  • Python
  • Java / Spring Boot
  • FastAPI / gRPC
  • REST Architecture
  • DSPy / LangChain
  • Dagster

Infrastructure & Cloud

  • Apache Kafka / Pub/Sub
  • GCP (Spanner, BigQuery)
  • Apache Beam / Dataflow
  • Docker / CI-CD
  • Grafana / OpenTelemetry
  • Uptrace

Specialised

  • Vector Databases (Qdrant)
  • Relational & Columnar DBs
  • MongoDB
  • LLM & Embedding Servers
  • FreeRTOS / C
  • BLE Protocol Stack
Education

PDEU

B.E. in Computer Science

2018 — 2022 · Gandhinagar

Pandit Deendayal Energy University — graduated with a strong academic record, building foundations in systems programming, algorithms, and embedded electronics that directly inform production engineering work today.

CGPA 9.6 / 10