Small team, big targets?Good. I know where to start.

Paid campaigns need a reason to reach a particular audience with a particular message. I built an AI workflow that combines CRM loss history with buyer conversations to define win-back audiences and campaign hypotheses. It turns those findings into message recommendations and campaign plans, with audience exports ready for the marketing team. The team used the system in paid-channel campaigns that brought prospects back into the pipeline.

Useful market conversations are easy to miss in a large stream of Reddit posts. I built a research and response desk that selects relevant discussions and explains where an expert could contribute. It produces short briefings with an angle, context and boundaries, then brings selected opportunities into Slack. The delivered work included the research, participation plan and working desk; sustained posting had not started.

Email metrics can reveal a problem without explaining its cause or the next useful action. I began a diagnostic workflow that turns an issue into an evidence-backed case and a proposed test. The early build investigated one audience-related issue and shaped the next version around decisions rather than another metrics dashboard. It remained a prototype, with broader investigation and action tracking still to develop.

A journalist’s request is useful only when the audience fits and someone can answer it credibly. I built a workflow that collects requests, checks those conditions and presents the relevant opportunities with deadlines and source context. I calibrated the selection through manual review and added Slack updates for new matches. The tool prepared opportunities for an expert to review and answer.

Supplier calls leave teams comparing proposals, notes and impressions from different conversations. I built an AI-assisted workflow that brings transcripts and proposals together with the team’s requirements. It organizes scores, strengths, weaknesses and price considerations in a comparison dashboard. The result helps people inspect the tradeoffs and explain a recommendation to management.

Early sales contacts needed a reason to stay engaged between conversations. I helped build a Telegram content bot with the marketing team, using sports news and AI-assisted content for a news group. Sales invited cold contacts into that group during their first interactions. It became a repeatable way to keep relevant content in front of those prospects.

Colleagues needed to see how AI could fit the work already on their desks. I designed and delivered two practical training sessions for go-to-market teams across LSports. We worked through recurring-task assistants, AI-assisted research and prototypes that people could discuss with design and web teams. The sessions gave colleagues methods they could adapt to their own workflows.

The next useful webinar speaker may already be somewhere in the CRM. I built a tool that combines those relationships with public profile information and available account context. It creates an evidence-backed shortlist that people can inspect by fit, expertise and location. The delivered result was a speaker library prepared for event planning.

A recorded webinar holds expert material that is difficult to reuse from a video alone. I built a workflow that identifies speakers, preserves attributable quotes and connects the content with CRM context. It organizes the material into a reusable library and prepares article drafts for human review. I manually checked reliability and worked through the field mapping needed to connect the sources.

I built a tool to make product videos easier to produce and revise, and turn long recordings into short promotional clips. Producing a few product videos had taken the team more than a week, with even small edits adding work. I built a workflow that renders scenes from code, so I could control the output and make precise changes. It also cuts webinar and summit recordings into short promos, with a person choosing from five versions.

An expert speaks differently in an article, a LinkedIn post and a live talk. I built format-specific writing guides and an AI workflow from an expert’s own material. The process separates product knowledge from voice and checks whether the output follows the right format. Reviewers rated the writing highly in testing, giving me feedback to refine the guides and the evaluation.