Rajan Jasani
Software Developer
2+ projects
Python
TypeScript
Node JS
SQL
JavaScript
Data Pipeline & Feature Engineering: 4/10
Active 10 days ago
Invite to interview
Download CVCV
Overview
Technical skills
Timeline
Roles
Overview
Frontend UI engineer specializing in high-fidelity, motion-driven web interfaces and editorial design systems. The strongest proven skill is custom SVG geometry and animation logic used to drive the living contour system, as implemented in lib/contour-shapes.ts and lib/contour-scenes.ts. Public code does not demonstrate complex server-side systems, advanced async state machines (cancellation/optimistic updates), or comprehensive automated test coverage.
Phone
Technical skills
Languages
5
Python
TypeScript
Node JS
SQL
JavaScript
Python
7
Django
FastAPI
Scrapy
Django REST Framework
SQLAlchemy
Pydantic
Flask
Databases
7
PostgreSQL
Redis
ElasticSearch
Apache Kafka
MySQL
pgvector
Pinecone
Frontend
4
Next.js
React.js
Tailwind CSS
Framer Motion
DevOps
9
AWS
Amazon S3
Amazon CloudWatch
Amazon EC2
Amazon ECS
AWS Lambda
Rest API
AWS Step Functions
GitHub
AI/ML
10
Embeddings
LLM
AWS Bedrock
Hugging Face
LangGraph
OpenAI
Anthropic
Semantic Search
LangChain
Whisper
Other
16
Playwright
Pytest
Docker
Git
Linux
Vite
PostCSS
CloudFormation
AI Agents
RAG
WebSockets
Prompt Engineering
IAM
Multimodal AI
SOAP
Semantic Search
Timeline
Software Engineer L2
•
Middle
Crest Infosystems Pvt. Ltd.
•
Full-Time
Built production backend and REST APIs for knowledge and learning platforms at multi-thousand weekly user scale using Python, FastAPI, Django, React, PostgreSQL, Redis, and AWS. Implemented ingestion pipelines for large volumes of RSS content with automated extraction and embedding-based indexing. Developed a curriculum-aware AI learning experience using vector search and asynchronous workflows for streaming responses and content evaluation.
Python
FastAPI
Django
PostgreSQL
Redis
React.js
AWS
Scrapy
Playwright
Embeddings
Semantic Search
Gujarat Technological University
Bachelor's Degree •
Computer Engineering
Backend Developer Intern
•
Junior
Crest Infosystems Pvt. Ltd.
•
Internship
Implemented REST APIs and backend business logic with Python, Django, and Django REST Framework, including database schemas and data validation/serialization. Worked with PostgreSQL to support API features and structured request/response handling for the learning platform services.
Python
Django
Django REST Framework
PostgreSQL
Middle AI/ML Engineer
Confidence: Medium LLM Engineer
Backend-focused LLM/RAG engineer at a Middle level with a strength for building production-ready retrieval and ingestion APIs. The strongest proven skill is designing resilient ingestion and serving endpoints with concrete artifacts in the ingestion and chat endpoints (backend/api/routes.py) that include size-limited uploads, batch handling, queue enqueueing, and SSE streaming for chat. There is little to no public evidence of custom model training, experiment tracking, performance optimization for GPU workloads, or production deployment manifests and monitoring.
Model Architecture & Training
How well models are designed and trained
Not evidenced in public code
Data Pipeline & Feature Engineering
4/10
How data is prepared for models
Practical ingestion and data-handling code is present with file size enforcement, batching, DB persistence and enqueueing for background processing; this shows real ingestion pipeline engineering though not a full ETL system with provenance/augmentation tooling.
Evidence
backend/api/routes.py: ingest_file (file size checks, disk write, DB create, enqueue_ingestion_task)
backend/api/routes.py: ingest_files_batch (batch loop, per-file cleanup, results aggregation)
Experimentation & Evaluation
How results are measured and tested
Not evidenced in public code
MLOps & Deployment
3/10
How models are shipped to production
Serving and operational concerns are addressed via FastAPI endpoints, SSE streaming for chat, background task enqueueing and a worker pattern; evidence of MLOps practices such as model versioning, monitoring, or deployment manifests is not present in the reviewed files.
Evidence
backend/api/routes.py: chat (Server-Sent Events streaming via StreamingResponse and stream_chat_response)
backend/api/routes.py: enqueue_ingestion_task usage (push to Redis for background workers)
Computational Efficiency
1/10
How efficiently computing resources are used
Some resource-conscious choices appear (streaming responses, chunked file copy with enforced size limits) but no evidence of GPU/batch optimization, quantization, or explicit performance profiling in the inspected human-authored code.
Evidence
backend/api/routes.py: _copy_uploadfile_limited (chunked read to avoid memory blowup)
backend/api/routes.py: StreamingResponse usage for incremental response streaming
Research Depth & Innovation
1/10
Depth of research and new ideas
There is design-level awareness of RAG workflows and hybrid retrieval integration calls, but no novel research artifacts, custom layers, or reproduced paper implementations in the reviewed human-authored code.
Evidence
backend/api/routes.py: references to hybrid_search and semantic_mmr_rerank indicating hybrid RAG design
backend/retrieval/fusion.rrf.py (module present indicating reciprocal-rank fusion approach)
Expertise
RAG• Middle
AI / LLM Engineering (Agents)• Middle
Technologies
Python• since 2023 • Middle
SQL
MySQL
PostgreSQL• since 2024
Redis• since 2024
LangGraph
LangChain• since 2026
pgvector
Pinecone
SQLAlchemy
CloudFormation
WebSockets
Embeddings• since 2024
Prompt Engineering
Scrapy• since 2024
Multimodal AI
AI Agents
AWS Bedrock
Git
AWS• since 2024
Docker
ElasticSearch
LLM
RAG
Whisper
Pydantic
Django REST Framework• since 2024
AWS Lambda
Amazon EC2
Semantic Search
OpenAI
Anthropic
Hugging Face
GitHub
Amazon S3
IAM
Amazon ECS
Amazon CloudWatch
AWS Step Functions
Semantic Search• since 2024
Linux
SOAP
Recommendations
- Build and harden RAG retrieval services and production LLM APIs that include streaming responses, citation handling and background ingestion pipelines.
- Implement end-to-end ingestion/ETL and vector indexing systems with explicit provenance, validation and data-augmentation steps to strengthen the data pipeline offering.
- Add experiment tracking and evaluation suites (W&B/MLflow or test harness) plus reproducible training/fine-tuning scripts if moving into model-development work.
- Focus on MLOps additions: deployment manifests, CI/CD for model/service releases, monitoring and drift detection for retrieval results.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Middle Backend Developer
Confidence: High Generalist
A middle-level software engineer focused on TypeScript and performant client-side rendering with a strength in algorithmic geometry and SVG path processing. The strongest proven skill is deterministic SVG path sampling and performance-minded rendering as implemented in lib/contour-shapes.ts (samplePath caching, curve flattening and discretization). The repository does not contain backend service implementations, database migrations, API surface code or observability/ops tooling to demonstrate production backend system design.
API Design
1/10
How well APIs are designed
No implemented API surface or API design patterns in the human-authored code; only declarative mentions of REST and backend technologies in portfolio metadata.
Evidence
PortfolioV3/lib/portfolio.ts: skillGroups includes 'REST APIs' and featuredProjects reference FastAPI
Data Layer & Database
1/10
Working with databases
No database code, migration history, or hand-tuned queries present in the analyzed human-authored files; databases only appear as declarative technology mentions.
Evidence
PortfolioV3/lib/portfolio.ts: featuredProjects and skillGroups list 'PostgreSQL' and 'pgvector' but no schema or migration artifacts
Scalability & Performance
3/10
Handling load and speed
Client-side performance and computational concerns are addressed with concrete tactics such as curve flattening, sampling discretization and a bounded cache for sampled paths, but there are no server-side scalability artifacts.
Evidence
PortfolioV3/lib/contour-shapes.ts: samplePath implements curve flattening, discretization and sampledPaths cache with eviction
PortfolioV3/lib/contour-journey.ts: entranceTrail and projectFrame compute fixed-size samples to avoid runtime geometry overhead
System Architecture
3/10
Overall system structure
Clear modular separation of concerns across focused utility modules for shapes, scenes and journeys shows deliberate organization at the library level, but no service decomposition, inter-service contracts or distributed-architecture design.
Evidence
PortfolioV3/lib/contour-shapes.ts, contour-scenes.ts, contour-journey.ts: separated modules with exported types and functions
PortfolioV3/lib/portfolio.ts: structured data exports for projects, skill groups and systems
Security & Auth
1/10
Protecting data and access
No implemented authentication, input validation, secrets management or boundary-level security controls in the human-authored code; security topics exist only as technology mentions.
Evidence
PortfolioV3/lib/portfolio.ts: skillGroups lists 'Cognito' and 'AWS' but no implementation or secrets handling code
Reliability & Observability
2/10
Stability and monitoring
Basic defensive coding and cache management are present (error on unsupported path command, bounded cache), but there is no structured logging, metrics, retry/backoff, timeouts or graceful-shutdown patterns.
Evidence
PortfolioV3/lib/contour-shapes.ts: throws new Error for unsupported path commands
PortfolioV3/lib/contour-shapes.ts: sampledPaths Map with eviction when size > 256
Expertise
Node.js• Middle
Backend AI & LLM• Junior
Industries
Education• Middle
Health Care• Middle
Technologies
Node JS• Middle
Rest API
Flask
FastAPI• since 2024
Django• since 2024
Apache Kafka
Recommendations
- Develop performant front-end visualization and animation libraries that require deterministic geometry sampling and caching logic.
- Extract and publish utility modules for SVG path sampling and smoothing for reuse in other web projects or design systems.
- Lead client-side performance work where deterministic rendering and low-latency drawing matter, for example interactive dashboards or media pipelines.
- Collaborate with backend engineers to expose small, well-scoped services that deliver precomputed geometry or server-side image rendering where needed
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Middle Frontend Developer
Confidence: Medium UI Engineer
Frontend UI engineer specializing in high-fidelity, motion-driven web interfaces and editorial design systems. The strongest proven skill is custom SVG geometry and animation logic used to drive the living contour system, as implemented in lib/contour-shapes.ts and lib/contour-scenes.ts. Public code does not demonstrate complex server-side systems, advanced async state machines (cancellation/optimistic updates), or comprehensive automated test coverage.
UI Component Architecture
6/10
How interface parts are built
Component and library structure shows deliberate separation of concerns, reusable motion/shape logic and custom UI primitives (Reveal, HeroMotion, contour libraries).
Evidence
StudioLocusV1/app/components/Reveal.js
StudioLocusV1/app/components/HeroMotion.js
PortfolioV3/lib/contour-scenes.ts
Responsive & Cross-browser
7/10
Works on all screens and browsers
Comprehensive responsive strategy with clamp(), fluid gutters, many media queries, logical sizing and responsive-fixes; images use next/image sizes and priority hints.
Evidence
PortfolioV3/app/portfolio.css
StudioLocusV1/app/globals.css
StudioLocusV1/app/responsive-fixes.css
Performance Optimization
6/10
Speed of the interface
Measured performance-minded patterns present: SVG curve flattening and caching, containment and reduced-motion fallbacks, image optimizations via next/image; no benchmark artifacts found.
Evidence
PortfolioV3/lib/contour-shapes.ts: samplePath implements curve flattening and sampledPaths cache with eviction logic
PortfolioV3/app/contour.css: contain: strict and prefers-reduced-motion rules
StudioLocusV1/app/page.js: next/image usage with sizes and priority for critical images
Accessibility & Semantics
6/10
Usable for everyone
Accessible signals integrated: skip-link, focus-visible outline, alt text and aria-labels on links, and prefers-reduced-motion support; evidence of conscious a11y choices rather than blind kit copy.
Evidence
PortfolioV3/app/portfolio.css: :focus-visible outline and .skip-link implementation
StudioLocusV1/app/page.js: aria-label on project links and consistent alt text usage
PortfolioV3/app/contour.css: prefers-reduced-motion handling and hiding decorative elements for print
State Management & Data Flow
2/10
Managing data in the app
Minimal client-side state complexity; server-side page generation and structured data are used, but there is no evidence of advanced async state discipline, optimistic updates, request cancellation or state machines.
Evidence
StudioLocusV1/app/projects/[slug]/page.js: generateStaticParams and generateMetadata server functions
PortfolioV3/lib/portfolio.ts: structured project data and systems metadata used to render pages
UX & Visual Polish
7/10
Look and feel quality
High visual polish with many microinteractions, tuned motion curves, hover/press states and reduced-motion fallbacks; strong editorial typography and layout discipline.
Evidence
PortfolioV3/app/portfolio.css: microinteractions, hover transforms and layered composition
PortfolioV3/lib/contour-scenes.ts and lib/contour-shapes.ts: motion math, deformTransition and actorScene for expressive animations
StudioLocusV1/app/globals.css: brand curtain, responsive typographic system and reveal patterns
Expertise
React• Middle
HTML & CSS• Middle
Modern Web Frameworks• Middle
Industries
Education• Middle
Health Care• Middle
Software• Middle
Technologies
JavaScript
TypeScript• Middle
Tailwind CSS
Next.js• since 2025 • 2 projects
React.js• since 2024
Vite
PostCSS
Framer Motion
Recommendations
- Lead development of design-forward marketing sites and portfolio experiences where animation and typographic control are key.
- Build reusable motion-driven component libraries and interactive SVG components for product and brand teams.
- Implement advanced frontend tooling projects that require performance tuning, image optimization and accessibility-first motion fallbacks.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories: