Large companies already have customers, revenue, and problems worth solving. In this conversation, Alephic forward-deployed engineer Josh Neland explains how small teams can use AI to test solutions quickly and put working tools into people’s hands.
Josh’s approach starts with a business goal and a bottleneck. Build something people can try, learn from their reactions, and use that experience to decide what comes next. He also explains why product judgment, engineering, and human review belong close together.
What stood out
Show people something they can use. Josh describes testing ideas with clients during discovery and getting prototypes in front of users in the first week. Early feedback helps the team find useful work before committing to a larger build.
Find the work that holds other work up. A routine service or repetitive task can be a good starting point when it slows the rest of the business. Improving it creates room for people to do more valuable work and helps the team understand the surrounding systems.
Keep the context close to the code. Small teams that combine product and engineering skills can carry what they learn from users directly into the build. Josh describes how fewer handoffs reduce coordination and preserve the meaning behind a request.
Make reviews short and frequent. Josh suggests starting with outputs a person can review in about two minutes. Those reviews expose incorrect assumptions and bring business knowledge from people’s heads into the workflow.
In this episode
00:00 Big teams, faster work
The question for the conversation: how can large companies use AI to shorten the path from a business problem to a useful tool?
00:37 Meet Josh Neland
Josh introduces his background in engineering, product management, and machine learning, and explains why he wants to stay close to building.
02:43 Build working tools in week one
How off-the-shelf AI capabilities and faster software development change discovery, prototyping, and the cost of finding out whether an idea is useful.
07:37 Start with the bottlenecks
Why improving routine work can free up capacity, reveal how the business operates, and create a foundation for more ambitious projects.
11:11 Product management meets engineering
Josh describes small teams that combine product judgment and technical delivery, with less information lost between the customer conversation and the code.
15:26 Revenge of the managers
How experience in strategy, resource allocation, and management can become useful again in hands-on engineering when AI expands what one person can deliver.
17:28 Breaking old work patterns
The challenge of changing inherited processes and organizational habits, and why small, visible improvements can help people experience a different way of working.
23:54 Find the right AI project
Start discovery with the team’s goals, the obstacles to reaching them, and what people have already tried. That connects a possible project to a business outcome.
25:39 Start with three experiments
Josh recommends trying AI on three real tasks. Keep the first outputs small enough to review quickly, and use the results to learn what help you need.
31:25 People check the work
Human review supplies judgment and missing context. Different stakeholders can use the same word differently, so capturing their perspectives matters to the finished tool.
37:56 How small teams build
How complementary skills, product fit, and regular code review help a small delivery team work together and learn from each other.
40:38 The enterprise AI advantage
Josh argues that established companies can combine existing customer needs and investment capacity with faster experiments to find value in their own operations.
44:54 Make the tools useful
A project example shows how early prototypes and user feedback reshape priorities. Ideas can return later as the team proves value and learns where to expand.
50:23 Working with Alephic
Bring a problem that matters to the business and involve the people doing the work from the start, so the team can build around their actual needs.
Published on YouTube October 9, 2026. Episode length: 52:29.
