AI Engineer with 1+ years in Machine Learning & Distributed Systems
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Highly motivated Computer Science and Engineering student with 1.0 years of experience in AI/ML development and system-level programming. Proven ability to design and optimize complex systems, from RAG pipelines and custom compilers to distributed key-value stores. Proficient in Go, Python, C++, and various ML/AI frameworks, with a strong foundation in data structures, algorithms, and system design.
Rajiv Gandhi Institute of Petroleum Technology
Bachelor of Technology · Computer Science and Engineering
August 1, 2023 – June 30, 2027
LFDT Hiero SDK
Open Source Contributor (C++, Systems)
October 1, 2025 – Present
India
AmberFlux EdgeAI Private Limited
AI Developer Intern
June 1, 2025 – August 1, 2025
India
Voice Agent Runtime
January 1, 2023 – June 1, 2026
Built a real-time voice agent achieving sub-1.1s end-to-end latency (speech in to first audio byte) on 4GB VRAM hardware by streaming each sentence to TTS as it completes, instead of buffering the full LLM response. Cut LLM time-to-first-token from 1.5s to 0.5s by profiling and removing cold-start and language-autodetection overhead in the speech-to-text stage. Designed a hardware-constrained inference split (STT/TTS on CPU, LLM on GPU at 2.2GB VRAM via Ollama), enabling concurrent multi-stage inference within a 4GB VRAM budget with zero GPU contention.
View ProjectBlan Compiler
January 1, 2023 – June 1, 2026
Built a C++17 Turing-complete interpreted language with a custom lexer and recursive-descent parser, and a memory-safe AST using std::unique_ptr and std::variant for dynamic type safety. Implemented a tree-walking evaluator supporting short-circuit logic and nested control flow, then extended the language into a deployable platform with a concurrent Go backend, JWT auth, snippet persistence, and Dockerized deployment.
View ProjectStrataKV
January 1, 2023 – June 1, 2026
Built an embedded Log-Structured Merge (LSM)-inspired key-value store in Go with WAL-backed durability, immutable SSTables, tombstone-based deletion, and background compaction. Sustained 330K+ operations on an append-only MemTable + SSTable architecture, reclaiming 80% disk space and reducing storage fragmentation from 1945 files to 2 files during compaction. Reduced negative-lookup latency to 163µs by integrating Bloom Filters to eliminate unnecessary disk I/O, while sustaining WAL-backed recovery of 2GB workloads in 632ms. Deployed StrataKV as the primary execution-cache layer for the Blan Cloud Engine, using SHA-256 request hashing for O(1) cache lookups, removing the need for an external Redis dependency.
View ProjectCompleted ML course
1Stop.ai in collaboration with IIT Roorkee
June 1, 2026 – Present
AI Developer Intern
AmberFlux EdgeAI Private Limited
June 1, 2025 – Present
Cultural Fit Analysis
The candidate's personal projects showcase a strong initiative and passion for building complex systems from scratch, which aligns well with an innovative and problem-solving culture. Their open-source contribution and internship experience demonstrate an ability to work within existing frameworks and contribute to team goals. The breadth of technologies and domains explored (AI, compilers, distributed systems) suggests adaptability and a continuous learning mindset, which are positive indicators for cultural fit in a dynamic technical environment.
Soft Skills & Operational Fit
The candidate's project descriptions highlight strong problem-solving skills, particularly in optimizing performance and resource utilization. Their open-source contribution demonstrates collaboration and adherence to best practices (decoupling protocols, strongly typed parsers). The detailed explanations of technical challenges and solutions suggest good communication of complex ideas. The diverse range of projects (AI agent, compiler, KV store) indicates a proactive and self-driven learning attitude.