Full Details
<h2>About the Company</h2><p><strong>Semifront Technologies Private Limited</strong> is a deep-tech VLSI startup based in Kolkata, building a scalable PCIe Controller IP spanning Gen1 through Gen5, with a focus on high-performance digital design and protocol-level implementation. The company is part of the NASSCOM DeepTech Launchpad initiative and was incubated under FABCI (Fabless Chip Design Incubator) at IIT Hyderabad, supported by the Ministry of Electronics & Information Technology (MeitY), Government of India.</p><p>Semifront engages in advanced architectural discussions with faculty members from leading IITs in semiconductor design domains, and maintains established business and marketing connections across the Bay Area (USA), Europe, and the Middle East to support global outreach and commercialization. Being a small, IP-licensing semiconductor company, confidentiality is non-negotiable — you will design AI workflows that follow strict rules about which data may go to which tools.</p><h2>Role Overview</h2><p>This is not a generic AI automation role. You will sit at the intersection of AI and silicon design — building AI systems that help engineers verify a PCIe controller IP, one of the most complex protocols in the industry. You will be the person who defines how AI is used across the company: in engineering, in HR and marketing, and in the infrastructure everything runs on. You own the work end to end, from the first conversation to the tool people rely on every day. You will work in person from the Kolkata office and report directly to the Founder.</p><h2>Key Responsibilities</h2><h3>1. AI for the VLSI Front-End Team</h3><ul><li>Sit with the design and verification team to understand their debug and regression workflows, and identify what can be automated.</li><li>Design the automation flow using AI agent skills, LLM pipelines, scripts, or whatever medium fits the problem best.</li><li>Build AI-assisted regression flows that run, monitor and summarise regressions and triage failures to a likely root cause.</li><li>Build AI-assisted debug tools: log and simulation-output analysis, lint/error summarisation, and debug guidance.</li><li>Build verification support tools: checklist extraction from the PCIe specification, coverage-gap analysis, and testbench/collateral assistance.</li><li>Test every flow with the team before rollout, with clear quality checks so engineers can trust the output.</li><li>Measure impact (time saved, faster debug) and keep iterating after deployment.</li></ul><h3>2. AI Automation for HR and Marketing</h3><ul><li>Work with HR and Marketing to identify repetitive processes and automate them — candidate screening support, onboarding workflows, document and report generation, content drafting, campaign and lead tracking.</li><li>Set up and connect automation platforms and AI tools with the systems these teams already use.</li><li>Debug, maintain and improve these automations so they keep running reliably.</li></ul><h3>3. Systems Administration</h3><ul><li>Handle day-to-day IT support for laptops and desktops (Windows/Linux) — setup, upgrades and troubleshooting.</li><li>Maintain Linux servers and user accounts, NFS shares and Git repositories for the engineering team.</li><li>Support the regression compute setup and job scheduling (e.g., Slurm).</li><li>Handle network, internet and VPN issues, and coordinate with ISPs and service vendors.</li><li>Install and test tools for engineering teams and keep system documentation up to date.</li></ul><h2>Tech Stack</h2><ul><li><strong>Languages & Scripting:</strong> Python (primary), Perl, Shell, Make</li><li><strong>AI / LLM:</strong> LLM APIs (Anthropic, OpenAI or similar), AI agents and agent skills, prompt/LLM pipelines, MCP tool integrations</li><li><strong>Automation Platforms:</strong> Zapier, Make, n8n, or custom-built orchestration</li><li><strong>Systems:</strong> Linux server administration (users, permissions, NFS, networking, VPN), Windows, Git, Slurm job scheduling</li><li><strong>VLSI / Domain Tooling:</strong> Verilog/SystemVerilog (basic understanding), verification and regression flows, simulators, lint tools such as Verilator, PCIe protocol specifications</li></ul><h2>What We Are Looking For</h2><ul><li>Diploma in IT/Computer Science or a Bachelor's degree (mandatory); B.Tech/BE in IT/CSE or any relevant stream (preferred).</li><li>3+ years of experience, or an exceptional fresher — hands-on work building AI/LLM-based automation counts, intensive recent project work included, not just job titles.</li><li>Strong Python; Perl, Shell or Make scripting is a plus.</li><li>Practical experience with LLM APIs (Anthropic, OpenAI or similar), AI agents, and automation platforms (Zapier, Make, n8n or custom-built).</li><li>Working Linux administration skills: users, permissions, NFS, basic networking and troubleshooting.</li><li>Ability to talk to engineers, HR and marketing alike, understand their problems, and turn them into working solutions.</li><li>A habit of testing thoroughly and measuring whether an automation actually saves time.</li></ul><h2>Nice to Have</h2><ul><li>Basic understanding of Verilog/SystemVerilog and verification concepts — the VLSI domain will be taught.</li><li>Experience in a semiconductor, EDA or VLSI environment.</li><li>Familiarity with simulators, lint tools (e.g., Verilator) and regression flows.</li><li>Experience building agent skills, MCP tool integrations or internal developer tools.</li></ul><h2>How to Apply</h2><p>Fill in the application form on the company website and include a link to something you have built with AI — an agent, an automation, a tool, or a short write-up. That matters more than your resume format.</p><h2>Location & Work Mode</h2><p>Kolkata, West Bengal — full-time, work from office, in person.</p>