THE
RÉSUMÉ.

AI infrastructure engineer and applied AI strategist building local-first systems for coding agents and decision support. I work across Python, Java, MCP servers, SQLite memory, hybrid retrieval, multi-agent simulation, market intelligence, predictive analysis, and full-stack delivery. My public work includes EpisodAI, Episoda Core MCP, ALPURIS OS, and a personal Sentinel / UWOF local AI lab.

Ataraxia Pharma Pvt. Ltd. | Applied AI, Simulation & Analytics | 2026

• Developed AI-assisted research and scenario-analysis workflows for wellness product planning, combining structured market inputs, synthetic consumer cohorts, and comparative evidence reviews.
• Worked with a 1,000,000-row Monte Carlo simulation dataset in SQLite; analyzed pricing, channel commissions, landed costs, net selling price, and margin variables through feasibility and profitability filters.
• Produced a ranked 1,000-scenario CSV export and an HTML simulation explorer to support inspection of model assumptions, trade-offs, and shortlisted scenarios.
• Applied predictive scenario analysis to acquisition cost, ROAS, conversion, channel mix, and unit economics; distinguished synthetic outputs and forecast assumptions from observed commercial performance.
• Structured competitor comparisons and scientific-evidence audits into spreadsheets, dashboards, and technical briefs, documenting source evidence, limitations, and validation requirements.

EpisodAI | Creator / maintainer | 2025–Present

• Built a local MCP memory engine with SQLite WAL, vector and trigram retrieval, knowledge-graph workflows, cognitive decay, and Obsidian ingestion.
• Hardened the threaded HTTP bridge for concurrent requests and documented reproduction commands instead of relying on unverified benchmark claims.

Episoda Core MCP | Creator / maintainer | 2025–Present

• Built a zero-dependency Python MCP server using SQLite FTS5/BM25 for offline, persistent recall by coding agents.
• Designed a small proof path—save, search, stats, and tests—to keep the system auditable and easy to adopt.

ALPURIS OS | Creator / maintainer | 2025–Present

• Built a local multi-agent city simulation with autonomous citizens, sector activity, civic governance, an economy, Python services, and a 3D WebGL voxel interface.
• Added repo-relative configuration, portable workspace/database overrides, Docker/Compose support, and a reproducible demo path.

Sentinel / UWOF local systems lab | Personal research project | 2025–Present

• Designed a local orchestration runner that routes between MLX model tiers, a web-scout path, and an OpenInterpreter/Ollama execution path.
• Maintains local model checkpoints and an MLX LoRA fine-tuning script for Apple Silicon experiments; runtime assumptions are documented in the source folder.

Independent Full-Stack Developer | 2025–2026

• Built web applications with React/JavaScript frontends and Spring Boot backends, following SOLID principles and clear API boundaries.
• Designed SQL schemas and Hibernate/JPA mappings; built REST endpoints with global exception handling and Postman validation.
• Applied manual QA and Selenium automation to improve regression coverage and release confidence.

Java Full Stack Development Certification | KodNest

Completed Java Full Stack Development training at KodNest; current location: Nashik, Maharashtra, India.

Java, Python, Spring Boot, JPA/Hibernate, React, SQL, MongoDB, REST APIs, DSA, manual QA, and Selenium automation.

Bachelor of Commerce in Computer Application | Rayalaseema University, Andhra Pradesh

Coursework included Python, Core Java, C/C++, R, DBMS, and enterprise application foundations.

GitHub: https://github.com/lalithbuilds · Project home: https://lalithbuilds.github.io/ · LinkedIn: https://www.linkedin.com/in/lalith-alpuri/